At half past one in the afternoon on Wednesday 19 August, 307,425 people had already registered for a free AI account. The volume was heavy enough that the ID verification system slowed, and officials pointed people to the Thang Rath app instead. By that evening the figure was around 500,000. Within two days it had passed one million.

The programme is called TH-AI Passport. It is the largest digital access project the Thai state has undertaken, and almost nobody in the Chiang Mai business community is discussing what it does to this market.

The basics

What you actually get

This is not a watered-down edition. The published list includes GPT 5.6, Claude Opus 5, Gemini 3.1 Pro, Grok 4.3, Sonar Deep Research, GPT Image 2, Nano Banana Pro, Seedance 2.0 and Lyria 3 Pro. Text, images, video, music, deep research. These are the paid tiers.

Bought separately on the open market, that stack runs to several thousand baht a month. The state is paying 24.70 baht per active user per month, having bought five million seats at once.

Two practical notes. The programme covers the consumer interfaces rather than developer API access, so anyone building software still pays their own way. And detailed usage limits have not yet been published, so plan on generous rather than unlimited.

Pathumma is the quiet upgrade

Every business in this city has watched a foreign AI model produce stiff, formal, slightly wrong Thai. It mangles Northern place names. It misses the register a customer expects in a Line message. It treats Thai as a translation exercise.

Pathumma LLM is different. It comes out of the national ThaiLLM project, it runs on servers inside Thailand, and it sits in the same dropdown as the American models.

For Line broadcasts, review replies, customer service and any copy aimed at a Thai audience, this is the change your staff notice on day one. The American names get the headlines. Pathumma does the daily work here.

Why 24.70 is the number to remember

For less than a coffee at a Nimman cafe, the guesthouse in Santitham that has never paid for a subscription now runs on the same tools as a Bangkok agency. So does the massage studio on Loi Kroh with a Facebook page and no website. So does the coffee roaster in Mae Rim whose product photography has always let the brand down.

The gap between well funded Chiang Mai businesses and everyone else was, in large part, a gap in what they could afford to buy. That gap closed in an afternoon.

The catch: 30 days, then coursework

Access is not unconditional. Users get roughly 30 days, after which continued access depends on completing courses from a library of more than 96, built on the UNESCO AI Competency Framework with material from Google for Education, Microsoft Elevate and OpenAI. Completion earns certificates and points, and points unlock the higher Innovator tier.

The design is deliberate. This is a skills programme with tools attached, not a tools giveaway with courses attached. It also means the registration figure and the eventual active user figure will separate, and the second number is the one that describes the real market.

The bit nobody has said out loud

Registration is for Thai nationals aged 15 and over. There is no expatriate route.

Inside a single Chiang Mai office, this creates something with no precedent. The Thai marketing coordinator has Claude Opus 5 through a free account. The foreign owner sitting three metres away is paying for it personally or working on a free tier.

Two things follow.

Capability shifts toward Thai staff. This is a genuine argument for delegating more, and for handing local teams work that foreign management has been holding onto.

Thai-owned competitors just upgraded their entire workforce at no cost. Any business here that depends on its foreign management for content, research and analysis is about to be outrun by companies where everybody got the tools.

Your staff will use this at work. Set the rules first.

The government has been explicit that SME entrepreneurs and workers are a target group. Your team will use it for work. Some already registered.

Here is where your own compliance obligation sits. Under Thailand’s Personal Data Protection Act, you remain the controller of your customers’ personal information regardless of which platform an employee pastes it into. That is true of every AI tool, every messaging app and every free service your staff use, and it has nothing to do with this programme in particular.

Know what the platform handles. It collects name, national ID number, occupation, income bracket, location, device information and the content of AI conversations including uploaded files. Conversation data is retained for up to a year. The ministry states that personal data and behavioural data are stored and processed in Thailand under the PDPA with cybersecurity safeguards in place, and that the contract prohibits providers from storing or reusing citizen data to train models. Model inference itself runs on international infrastructure, as it does with any global AI service.

A front desk staff member drafting a reply to a guest complaint moves that guest’s name, booking reference and grievance into a third party system. The programme advises users not to enter sensitive information, which is sound advice. Acting on it is your job, not the platform’s.

No business in Chiang Mai has a written policy for this, because until last week there was nothing to write one about.

Where advantage goes now

Producing a decent social post, product description or customer reply has stopped being a skill worth having. Five million people acquired it in a fortnight.

The revealing statistic is not about tools at all. Research released by AWS this year found 43 per cent of Thai businesses now use AI consistently, up from 32 per cent, while only 21 per cent have a formal AI strategy. Adoption is wide and shallow. Most businesses use AI to produce the same things faster, rather than to reconsider what they produce.

Three things become more valuable from here.

Verified information about your actual business. Ask any model about your restaurant and it assembles plausible sentences from whatever it can find. It cannot invent that the kitchen closes at nine, that the upstairs room seats 24, or that you have run the place since 2011. That exists only where you have published it clearly and consistently. Businesses that have done that work become the answer. Businesses that have not become the gap in the answer.

Your own data. Booking patterns, customer history, supplier terms, ten years of knowing which week in November goes quiet. None of that sits in any model. The businesses that pull ahead this year are the ones feeding their own information in, rather than asking a tool to guess at context it does not have.

Deciding what is worth making. Output is now close to free, so the market fills with competent, forgettable work produced by people who suddenly can rather than people who should. Choosing what to make was never a technology problem, and it has just become the entire game.

Five things to do this month

1. Write the policy. One page. What staff may and may not put into AI tools, all of them, not just this one. Customer names, contact details, bookings, health information, financial records, employee files, anything received in confidence. In Thai and English. Have people sign it.

2. Let your team register, then train them. Registering before 30 November grants the higher tier. Your Thai staff get a capability upgrade you do not pay for. Take it, and spend an hour explaining how you expect the tools to be used.

3. Check what AI already says about you. Ask three different models about your business. Where the answer is wrong, thin or missing, fix the underlying information on your website, your directory listings and your Google Business Profile. Everyone is about to publish more. Only businesses with verified data underneath get cited.

4. Move the work around. Push Thai-speaking staff toward the tasks where these tools give them real leverage. Move senior attention toward deciding what is worth producing at all.

5. Do not hire on certificates. Credentials from this programme start appearing on CVs by the fourth quarter. Give candidates a real task from your business instead. The certificate proves they finished a course. The task tells you whether they can think.

Two numbers to watch

Active users against registrations. The contract pays only for people who actually use the platform, so the active figure is the one that describes real adoption in this market. The gap becomes visible by October.

Course completion at day 31. Access continues for users who finish training. That threshold determines how many of the five million are still working with these tools in the fourth quarter, and therefore how much the competitive floor has genuinely moved.

Keep your own copy of everything

One piece of practical hygiene that applies to any platform you do not own. Keep your prompts, templates, workflows and outputs stored somewhere under your control. Programmes end, terms change, and a twelve-month access window closes in August 2027. The value your team builds over the next year should live in your business, not only inside somebody else’s login.

Chiang Mai has spent a decade with a quiet division running through it. The foreign population arrived with skills and subscriptions. The Thai-owned commercial base ran on Facebook, Line and word of mouth. That division shaped who competed with whom, and who never had to.

On 31 August, five million Thai citizens receive tools that cost everyone else a credit card to reach. The course requirements will thin the numbers considerably, and shallow use will limit what most people get out of it. Neither changes the floor this puts under every competitor in this city.

Businesses that treat this as free software will produce more content and win nothing. The ones that treat it as the moment tools stopped being the advantage, and move their effort to verified information, their own data, and better decisions about what to build, will finally get paid for having done the harder work all along.

CMBN Staff

chiangmaibusiness.net/

Ask any assistant for the opening hours of a small business in Chiang Mai and the answer arrives in a confident, complete sentence. Check that sentence against the business itself and the confidence turns out to be unearned more often than the fluency suggests. The systems producing these answers have no mechanism for signalling the difference between a fact retrieved and a fact assembled, so a guess and a certainty leave the model looking identical.

That gap has now been measured, and the measurement contains an uncomfortable finding for every small operator in this city.

Small businesses absorb the error rate

Searchable, an AI visibility platform, ran more than 13,000 prompts through ChatGPT, Perplexity and Gemini about real London companies, comparing small and medium enterprises against firms of 500 employees or more. Every answer was checked against verified records including Companies House filings and official company biographies. Across the full sample, 93% of companies had at least one basic fact hallucinated or missing from an assistant’s answer.

The distribution of that error is the part that matters for anyone running a business here.

Half of the SMEs in the study received at least one outright fabricated fact, against 32% of large companies, a rate 56% higher. Assistants confidently stated false information about SMEs at 5% against 2% for large brands. Key information went missing entirely on 6.3% of SME queries and 4.8% of large company queries. Brand name confusion, where a model attributes one company’s details to another, occurred at 4% for SMEs and 0.7% for large brands, a fivefold difference. Services were described accurately 87% of the time for SMEs and 93% of the time for large brands. Aggregated across every prompt, 11 in 100 questions about SMEs returned false or missing information against seven in 100 for large companies.

The three categories the models handled worst were company size, contact details and founding year. Those are the facts a customer needs before they walk through a door.

The study covered London rather than Chiang Mai, and the specific percentages do not transfer directly to a Thai market with a different language mix and a different indexed footprint. The mechanism transfers completely. Local Falcon’s analysis of the same data puts the cause plainly: even a business with an accurate, well-optimised website struggles to surface reliably against a competitor carrying substantial third-party validation across the web. Models fabricate where the public record is thin, and the public record is thinnest around businesses that nobody other than the business itself has ever written about.

Thailand is further into this than most markets

The assumption that AI search is a Western phenomenon arriving here on a delay does not survive contact with the adoption data.

DataReportal’s Digital 2026: Thailand report counted 67.8 million internet users at the end of 2025, putting online penetration at 94.7%, alongside 56.6 million social media identities equal to 79.1% of the population. Social identities grew by 7.5 million in a single year, a 15.2% increase ranking among the fastest in the Asia-Pacific region. The typical Thai internet user spends around 34 hours and 32 minutes online each week.

Microsoft’s country-level adoption data records Thai AI usage growing 36.2% between the first half of 2025 and the first quarter of 2026, the second largest increase globally behind South Korea and roughly double the United States rate of 19%. The same analysis notes that adoption in low and middle income countries is growing at more than four times the rate of the wealthiest nations.

Platform concentration here is unusual. Statcounter data from December 2025 gives ChatGPT 74% of Thai AI chatbot traffic, above its global average, with Gemini at 16%, Perplexity at 7%, Copilot at 2% and Claude under 1%. Thailand also remains one of the most Google-dominant search markets in the world, with Google holding approximately 99.56% of Thai search traffic in early 2026. A Chiang Mai business therefore faces a narrow and specific exposure: ChatGPT for conversational discovery, and Google AI Overviews for everything else.

Thai businesses have adopted the technology enthusiastically. The UOB Business Outlook Study 2026, surveying 265 Thai business owners and senior decision makers, found more than seven in 10 Thai SMEs implementing AI and more than eight in 10 using digital tools, both above the regional average. Very few of them have audited what the technology says about them.

The Thai language problem nobody is costing

There is a second exposure specific to this market, and it receives almost no attention in the local marketing conversation.

Large language models perform measurably worse in lower-resource languages, a pattern documented well enough in the research literature to have its own name. Work presented at EACL 2026 on hallucination detection across languages found that task accuracy drops sharply when models move from English to lower-resource languages, across factual recall, STEM and humanities domains. A parallel study on factual accuracy in English versus low-resource languages found models frequently perform better in English even on questions rooted in local context, with a higher tendency toward hallucination in the local language.

Thai carries additional structural difficulty. The language has no spaces between words, which complicates tokenisation, and Thai discourse norms soften definitive statements in ways that introduce apparent uncertainty into otherwise factual content. Dedicated Thai-adapted models exist, including Typhoon and the SEA-LION family, yet none of them powers the assistants Thai consumers actually use.

The practical consequence for a Chiang Mai business is a two-sided risk. A business documented only in Thai carries a thinner English record for models that reason best in English. A business documented only in English disappears from Thai-language queries entirely. Businesses serving both a Thai domestic market and an expatriate or tourist market need their core facts stated cleanly in both languages, in visible text, on surfaces that get crawled.

The confidence gap runs in both directions

Executives trust these systems more than the systems have earned. Workiva’s 2026 Midyear Executive Benchmark Survey, published in August, polled 2,272 finance, risk and sustainability professionals including 847 C-level executives, alongside 367 institutional investors. It found that 84% of executives are at least somewhat confident in the accuracy of AI output without human review, broken down as 39% somewhat confident and 45% very confident, while 26% report that internal audits caught AI errors which reached external audiences or board members. Only 11% believe their organisation’s data quality is sufficient for AI use, and 89% of the institutional investors surveyed are concerned about the accuracy of AI-generated information in disclosures, with 47% actively looking for such errors.

Consumers have moved the other way. Fractl, working with Search Engine Land, surveyed 1,008 United States consumers and 150 marketers in the second quarter of 2026, repeating questions asked a year earlier. The share finding AI more helpful than traditional search fell from 82% to 54% in twelve months, a 28-point drop. The group actively rating AI as less helpful grew from 3% to 17%, close to six times larger. Over the same period, 70% reported using AI for search more than they did the year before, and only 3% reported using it less.

Two details in that dataset complicate the standard demographic assumptions. Baby Boomers now find AI search more helpful (63%) than Gen Z does (47%), meaning the cohort with the most exposure is the most disillusioned. And the share of consumers saying heavy brand use of AI would reduce their trust in that brand roughly doubled from 20% to around 40%.

Local discovery has shifted faster than general search. BrightLocal’s Local Consumer Review Survey 2026, a representative panel of 1,002 United States adults published in February, found consumers using AI tools to find local business recommendations rising from 6% to 45% in a single year. That makes AI the third most-used local discovery channel behind Google and Facebook, ahead of Yelp and Tripadvisor. Google’s own share of local business discovery fell from 83% to 71% across the same period. BrightLocal’s supplemental AI trust report breaks the figure down further: ChatGPT was used by 31% of consumers for business recommendations, Google AI Mode by 23%, and 64% of consumers aged 30 to 44 have asked AI for a business recommendation against 24% of those aged 60 and over.

What people do after the answer appears

Yext surveyed 3,848 consumers globally in 2026 about local search behaviour, with an AI-user subgroup of roughly 1,630 respondents. The headline number reshapes how visibility should be understood: only 5% of consumers move directly from an AI answer to a purchase.

Everyone else verifies. After receiving an AI recommendation, 53% run a Google or Bing search to confirm or learn more, 49% visit the business website directly, 42% click through to the sources the AI cited, 28% look for reviews on Google or similar platforms, and 20% check social profiles. Those rates hold almost constant regardless of stated trust levels. In the same survey, 75% of AI users rate their trust in AI local recommendations at four or five out of five, and 59% say that trust has grown over the past year, and they verify anyway.

The Yext data also establishes a spending threshold that determines where this matters commercially. The median amount consumers globally are comfortable letting an AI spend without review is 25 United States dollars. Every commercially meaningful local decision, from a hotel booking to a dental appointment to a contractor quote, sits above that line.

What tips the decision once a consumer starts verifying is review signal. Yext found star ratings on third-party platforms ranking first among purchase influencers after an AI recommendation, review recency second, word of mouth from a trusted contact third, and review content and sentiment fourth. BrightLocal puts hard floors under those signals: 47% of consumers will not use a business with fewer than 20 reviews, and 31% will only use one rated 4.5 stars or higher, up from 17% a year earlier.

One counterweight belongs in the record, because the verification story is often told too neatly. GatherUp research reported by Search Engine Land found that 67% of consumers do not rigorously fact-check AI sources before choosing a local business. The two findings coexist because verification is uneven rather than absent. Consumers check something, and what they check is rarely the full source trail. A wrong answer that survives a casual glance at a business website still costs the booking.

The schema correction

The standard remedy sold against this problem is structured data. Add JSON-LD markup, define the business as a machine-readable entity, and AI systems read the facts directly instead of guessing. The 2026 evidence does not support that claim, and the gap between the claim and the evidence has become the most expensive misunderstanding in local marketing.

Ahrefs published a controlled study in May 2026, authored by Louise Linehan and Xibeijia Guan. The team began with an analysis of 6 million URLs which found that pages cited by AI were almost three times more likely to carry JSON-LD. Rather than publish that correlation as a finding, they tested it. They identified 1,885 pages that added JSON-LD for the first time between August 2025 and March 2026, matched each against control pages from other domains with similar pre-existing citation levels, and measured citations across Google AI Overviews, Google AI Mode and ChatGPT for 30 days either side of the change.

Google AI Mode rose 2.4% and ChatGPT rose 2.2%, both statistically indistinguishable from zero. Google AI Overviews fell 4.6%, a decline the authors report as statistically real while declining to attribute it to the markup itself. Four separate tests pointed the same way, including an event study checking for pre-existing divergence between the two groups.

The explanation for the original correlation is straightforward. Structured data lives on well-maintained, technically competent sites, and those sites also publish better content, earn more references and rank better in conventional search. The markup was riding the signal rather than creating it.

Ahrefs cross-referenced a separate experiment by searchVIU, which tested five major systems, ChatGPT, Claude, Perplexity, Gemini and Google AI Mode, for real-time schema use when fetching a page. None of them read it. During direct retrieval, every system extracted only visible HTML and ignored JSON-LD, Microdata and RDFa entirely.

Google states the same thing in its own documentation, which says plainly that no new machine-readable files or markup are required to appear in AI features. John Mueller has said structured data will not make a site rank better and that most sites see no visible change from adding it.

Three honest caveats belong alongside this, and any consultant presenting the study without them is misrepresenting it. Every page in the Ahrefs treated group already carried 100 or more AI citations before the change, so the study answers what schema does for pages already visible rather than what it does for a business with no footprint at all. The study pooled all schema types together, so type-specific effects remain unisolated. And Microsoft’s Fabrice Canel confirmed on record in March 2025 that Bing’s systems do use structured data, making Bing the one platform with a first-party confirmation.

Schema remains worthwhile as indexing infrastructure, entity resolution and rich-result eligibility. It is hygiene rather than a growth lever. Any agency in this city selling markup as an AI visibility product in 2026 is selling against the published evidence.

Where the citations actually come from

Omniscient Digital analysed more than 23,000 AI citations on branded queries. A brand’s own website accounted for 23% of them. Earned sources accounted for 48%, and commercial content published by other companies accounted for 30%, with rounding taking the total marginally over 100. Reviews and other forms of social proof accounted for 57% of citations on branded queries, and the researchers found that educational content wins during early discovery while social proof wins when buyers approach a decision.

Roughly 77% of the citation surface sits outside the domain a business controls. The same directional finding appears in independent work: a Search Engine Land analysis with Evertune covering roughly 25,000 of the most-cited URLs found that across nearly 400 million citations, 63% pointed to ranked listicles rather than brand homepages.

This is the Searchable finding approached from the opposite end. Chris Donnelly, Searchable’s co-founder, attributed the SME error rate directly to footprint, noting that platforms train on publicly available web data which skews toward larger brands. His recommended starting point was building third-party validation through directories, reviews, articles and online communities. Searchable also cites Similarweb data indicating that an AI recommendation drives roughly twice the traffic to a brand’s website within seven days compared to a competitor which was not recommended.

The two datasets converge on one conclusion. The asset determining how an assistant describes a business is not the business website, and it is not the markup on the business website. It is the density and consistency of independent references to that business across surfaces the model treats as corroboration.

Small channel, high intent

The commercial case for this work rests on quality of traffic rather than volume, and the honest version of that case includes both halves.

Multiple independent measurements agree that AI referral traffic converts far above conventional organic search. Semrush established the widely cited cross-industry baseline at 4.4 times the conversion rate of standard organic. Ahrefs found internally that AI-referred visitors made up 0.5% of sessions and drove 12.1% of signups, a 23-times differential representing the upper bound. Rocket Agency’s 18-month cross-industry study found ChatGPT visits converting at 5.1 times organic. The mechanism is intent: the model has already defined the problem, synthesised the options and produced a shortlist before the visitor arrives.

The counterweight is volume. Conductor’s 2026 benchmarks put AI referral traffic at approximately 1.08% of all website visits, despite year-on-year growth of 527%. A conversion premium applied to one per cent of traffic produces a fraction of the revenue impact of ordinary conversion applied to forty per cent of traffic.

The pressure on the older channel is what changes the calculation. BrightEdge recorded AI Overviews triggering on around 48% of tracked queries by February 2026, up from roughly 30% a year earlier, while Seer Interactive measured organic click-through rates falling 61% on queries where an AI Overview appears. Seer also found that brands cited in AI Overviews earn 35% more organic clicks and 91% more paid clicks than uncited brands on the same queries. The AI channel remains small in absolute terms. The channel it is eroding is not.

A note on the figures in circulation

Several numbers travel widely in this field without a traceable source, and an article about machine-generated misinformation has an obligation to hold its own figures to the standard it asks of the machines.

Claims that schema makes content 36% more likely to appear in AI summaries, that it produces 2.5 times higher citation rates, or that businesses lose 60% of visibility without it, do not resolve to published methodology. As of March 2026 there were no peer-reviewed studies on the effect of schema on AI search visibility. The 2024 Princeton and Georgia Tech GEO paper cited constantly in schema discussions tested in-content strategies such as adding citations and statistics, which worked, and did not test schema at all.

Two further figures in wide circulation deserve flagging. The claim that a specific high percentage of local SMEs lack a GEO plan has no identifiable primary source. The documented version comes from Centerfield research shared with Search Engine Land, which found 63% of marketers saying their company is not investing time or budget in GEO, only 9% with resources allocated to it, and 33% claiming a good or expert understanding of GEO against 72% for conventional SEO.

Where a figure below carries vendor research rather than independent study, the vendor is named. Readers can weigh the incentive accordingly.

The audit any owner can run this week

The practical sequence follows from the evidence rather than from the sales pitch, and the first step costs nothing but an hour.

Interrogate the assistants directly. Ask ChatGPT and Gemini the questions a customer asks. Opening hours. Address. Phone number. Services offered. Price range. Ask in Thai and in English, because the answers diverge. Record every response and mark each error.

Trace each error to its source. Wrong facts almost always come from a stale listing, an aggregator nobody claimed, or a page that states the information inside an image, a PDF or a booking widget. Retrieval systems read rendered text and ignore everything else.

State the facts in visible text. Opening hours, address, phone number, services and pricing belong in readable sentences on the website, in both languages where both markets matter. Dated first-party statements give retrieval engines something specific and recent to cite.

Reconcile every external listing against that source. Conflicting addresses across secondary platforms give a model competing candidates and no basis for choosing between them, which is precisely the condition under which fabrication occurs.

Check that AI crawlers are not blocked. Many sites block OAI-SearchBot, PerplexityBot, ClaudeBot and Google-Extended through default security settings without the owner knowing.

Build presence on the surfaces carrying the citation weight. Directories, review platforms, local publications and community references supply the corroboration owned channels cannot generate alone.

Re-run the audit quarterly. Corrections take weeks to months to propagate, and new stale listings appear continuously.

Where Golden Pages fits

Golden Pages was built on the assumption that a verified directory listing does two jobs at once, and the 2026 data confirms both.

The first job is corroboration. A verified, human-checked listing adds an independent reference to a business at exactly the layer where roughly 77% of AI citations originate, and where thin-footprint SMEs currently lose to larger competitors by a factor of five on brand confusion alone. Searchable’s own recommendation to businesses appearing incorrectly in AI answers was to build third-party validation through directories.

The second job is verification. Of the consumers who receive an AI recommendation, 42% click through to the cited sources and 53% run a confirming search. A listing agreeing with the business, the reviews and the assistant closes that loop. A listing contradicting any of them breaks it, and the consumer moves to a competitor whose record is consistent.

Membership of the Chiang Mai Business Network adds the layer no crawler manufactures, which is a record of businesses that other businesses in this city have actually dealt with. Peer endorsement was the strongest signal in local commerce before assistants arrived. A technology that fabricates facts about small companies at 5% has raised its value rather than reduced it.

The next era of local discovery belongs to businesses whose facts agree with themselves everywhere a machine reads and everywhere a customer checks.

CMBN Staff

chiangmaibusiness.net/

The short version, for owners who do not work in marketing

• Google now answers the question itself, in its own words, above the list of links. Most searches end there without a click.
• Search traffic to small sites has collapsed. Chartbeat data puts the drop at roughly 60% for small publishers over two years.
• In May 2026, Google published its own written instructions on what makes a business visible inside AI answers. The document is free, short, and written for non-specialists.
• There is no trick, no special code, and no agency shortcut. Google rewards material a competitor cannot copy: your own numbers, your own cases, your own experience.
• An AI answer names two or three businesses. Coming eighth in your category produces nothing, where it once produced enquiries.
• AI referrals remain small in volume. This is a shift in how customers decide, not yet a replacement traffic source.

Three things to do. Read Google’s own guide to generative AI features before you pay anyone to interpret it. Publish content only your business is able to produce. Measure how often your name appears in AI answers instead of counting visitors.

A search for “best boutique hotel Chiang Mai” now returns a paragraph before it returns a hotel. The paragraph names two or three properties, explains in a sentence each why they are worth booking, and places small citation links beside the text it drew from. Below the paragraph sits the older version of search, the ten ranked links that built an entire local marketing industry, and most people never scroll that far.

Two things happened in Google Search over the past three months, and the second matters more than the first. Google redesigned how its AI layers present the web, which was expected. Google also published official documentation explaining how to be visible inside those layers, which was not. Simon Moser set out the commercial consequences in Entrepreneur on 6 August. The documentation removes the last excuse available to any business owner, which is not knowing what the rules are.

In early May, Google updated AI Mode and AI Overviews, the AI-generated layers that sit above conventional results. The updates added suggested follow-up angles at the end of an answer, site previews when a user hovers over a link, prominence for news publications a user already subscribes to, and citations positioned next to the specific claims they support.

These read as interface adjustments, and they are not, because together they confirm that the synthesised answer rather than the ranked list is now the thing Google delivers. The user reads a response and then decides whether any onward click is necessary at all.

The traffic consequences are recorded. Chartbeat data covering thousands of sites shows that search referral traffic fell by roughly 60% for small publishers and 47% for mid-sized ones across two years, with large publishers losing 22%. The gap between those figures is the story. Businesses with a recognised name and a direct audience absorbed the change, and businesses that depended on being found absorbed the full force of it. Our quarterly digital audits find the same pattern across Chiang Mai, most visibly for operators whose enquiries once arrived through blog content and category pages.

In mid-May, Google released its first official guide to optimising for generative AI features. The central position is that optimising for AI search remains SEO, performed properly, rather than a separate discipline requiring separate tooling.

The guide is most useful for what it rules out. There is no special schema markup that unlocks AI visibility, and Google names llms.txt files, content chunking, and AI-specific rewriting as things site owners can ignore. Manufacturing brand mentions across unrelated websites is far weaker in effect than the agencies selling it suggest. What the documentation rewards consistently is content an AI system cannot assemble from anywhere else: original research, first-hand experience, subtopic depth, material that exists because a specific business did the work.

On 5 June, Google extended its long-standing advice on hiring SEO help with a new document on assessing third-party tools, services, and advice, naming generative engine optimisation and answer engine optimisation as service categories for the first time. Google states plainly that it does not evaluate or endorse third-party services, and it supplies vetting questions for owners to use directly. Does the provider cite official Google documentation. Is the AI optimisation advice consistent with what Google has published. Anyone in this city who has sat through a pitch promising guaranteed placement in AI answers now has a free instrument for testing it.

Two claims in the current coverage deserve more scepticism than they receive, and both cut against the urgency being sold around this subject.

The first is the idea that AI referrals are quietly replacing lost search traffic. They are not, at least not yet at any meaningful scale. Chatbot referrals grew sharply through 2025 and still account for less than 1% of publisher page views. The visitors who arrive from an AI recommendation do convert at a higher rate, because the comparison stage happened before they landed, and a higher conversion rate on a very small number is still a very small number.

The second is Google’s own position, which is that AI Overviews increase search usage overall and produce higher-quality clicks. Google has published data supporting this and has an obvious commercial interest in the conclusion. Both things are consistent with the publisher figures, because total search activity grows while the share of that activity reaching any individual small site falls.

The honest reading sits between the two. What has changed decisively is where the buying decision gets made. What has not changed is that traffic volume is no longer the number worth defending.

Dozens of Chiang Mai businesses compete for the same handful of commercial queries in every sector. Spa treatments, international schools, dental work, coworking, wedding venues, long-stay accommodation. In the ranked-list era, placing eighth in that field still produced enquiries, because visitors browsed the field and built their own shortlists.

That behaviour is ending, and the market structure changes with it. When an assistant returns three recommendations in conversational form, most users accept the three. There is no page two to venture onto. The recommendation is the shortlist, and the shortlist is the market.

The distance between being cited and not being cited is therefore not a matter of degree. A business that appears in AI answers accumulates advantage with every query in its category, and a business that does not appear is invisible to a growing share of the people actively looking to buy from it. Google’s new features create a few additional slots inside the answer through suggested angles and community perspectives, and every competitor in the category is contesting those same few slots.

Chiang Mai carries a specific disadvantage here. A large share of local business content is commodity content, produced cheaply, describing what any competent operator in the sector also offers. AI systems synthesise that material from a thousand interchangeable sources and cite none of them. The operators with genuine first-hand material, the vineyard that logs its own harvest data, the clinic that publishes treatment outcomes, the hotel that documents what guests actually ask for, hold an advantage they have not yet priced.

Run this test before you read any further

This takes 20 minutes and costs nothing.

1. Write down the five questions a customer asks immediately before buying from you. Use their words, not your marketing words.
2. Put each one into Google AI Mode, ChatGPT, and Gemini. Add “in Chiang Mai” where it applies, and run the Thai-language version as well if you serve Thai customers.
3. Record every business named in each answer. That list is your real competitive set, and it will differ from the one in your head.
4. Check what is cited beside the answer. Where the citation goes to a booking platform or a listings site rather than to any business website, the businesses in that category have handed the customer relationship to a third party.
5. Repeat the same five queries in 90 days. The direction of travel matters more than any single result.

Read Google’s optimisation guide before commissioning anyone. It is short and written for non-specialists, and it now functions as the benchmark against which every AI visibility proposal is measured. A vendor whose recommendations contradict it is selling something else.

Move content strategy from volume to depth. One piece of original research on your own sector outperforms 20 posts assembled from the same sources your competitors used. This is the single change with the clearest return, and it is also the change most local businesses resist, because it requires the business to have something to say.

Change what the dashboard measures. Raw traffic describes a shrinking game. Track how often the business is named in AI answers for the queries that produce revenue, and track the conversion rate of visitors arriving from AI referrals. Fewer clicks are doing considerably more work, and a dashboard built for 2023 hides that entirely.

Budgets are already moving. PMG, one of the largest independent marketing agencies in the world, advises clients to pilot generative engine optimisation at 1.5 to two times their existing search spend, according to Matt Allfrey, its head of SEO for EMEA. A Semrush study published in June 2026, covering 481 marketing professionals, found more respondents planning to invest in AI search optimisation than in conventional SEO, at 38% against 36%. In the same survey, 85% said AI has already changed how they approach search.

The businesses that will hold the recommendations in Chiang Mai 18 months from now are making one decision this quarter, which is to publish something a competitor cannot copy and to stop measuring the wrong number. Everyone else is reading traffic reports and waiting for figures that are not coming back.

If you run the 20-minute test above and your business does not appear, the next question is why, and that usually comes down to whether an AI system can reach your pages at all. CMBN Media Services runs a digital audit that checks exactly that: crawlability, structure, and whether your information is readable and attributable to you. Most Chiang Mai sites we audit fail before content is even part of the conversation. Email gee@chiangmaibusiness.net with your URL and we will run it and send you the findings.

CMBN Staff

chiangmaibusiness.net/

He took gold in Code Combat and silver in Code Mentum, two disciplines with distinct demands. Code Combat asks contestants to apply the building blocks of programming, variables, conditional logic, loops and functions, to construct instruction sets that solve assigned problems accurately and at speed. Code Mentum raises the bar to algorithm design, where the fastest and most efficient solution to a progressively harder task wins. Both run in Python under strict time limits, and both weigh analytical reasoning as heavily as syntax. The Olympiad awards its medals by performance band within each category rather than by finishing position, so a gold medal records that a standard was met outright, not that everyone else had a bad day.

The pipeline behind the medal

One medallist does not change an economy. The system that produced him is a different matter. Thailand has spent years pushing coding education into public and private schooling, and the national bet on AI capability, including the government’s TH-AI programmes and the ThaiLLM initiative, rests on whether that schooling produces people who can build things. A Thai student outperforming a global field in algorithm design at age 12 is early, concrete evidence that it does.

Chiang Mai has a direct stake in this. The city holds one of the deepest concentrations of tech education and training outside Bangkok, from international school computer science programmes to private coding academies serving Thai families and resident nomad households alike. Parents here weigh STEM provision when choosing schools, and every operator in that market now has a national proof point for the category they sell. The path from a Thai classroom to a podium in Rome has been walked, and it enters the sales conversation whether those operators raise it or not.

There is a hiring dimension as well. Software and digital services firms in Chiang Mai’s startup and tech sector routinely name talent depth as the constraint that caps their growth, and the strength of the school-age pipeline sets what the local hiring pool looks like in 2032 rather than 2026. A firm weighing a long-term engineering presence in Thailand against treating the country purely as a sales market now has one more data point on the side of building.

A pattern, not a headline

Nipun’s result sits inside a wider run of young Thai wins on international stages this year, across science awards and conservation prizes, in fields that map directly onto the industries Thailand says it wants. The generation now in Thai classrooms is competing globally and winning. For the schools, academies and employers of this city, that is the fact worth acting on.

Source: Khaosod English, 13 July 2026.

CMBN Staff

chiangmaibusiness.net/

For most small businesses in Chiang Mai, the immediate reaction will be: so what? Another government tech initiative, another acronym. But this one is different, and the reasons why matter directly to how you run your operations, market your services, and compete in an economy that is shifting fast.

The Gap That ThaiLLM Fills

Global AI models, including the dominant tools most Chiang Mai businesses already use, are trained overwhelmingly on English-language data. They understand Thai, but often imperfectly. Colloquialisms, regional dialect, formal administrative language, and the particular way Thai customers phrase questions and complaints all sit in a blind spot for Western foundation models.

ThaiLLM was built specifically to close that gap. It was trained on government documents, educational research, and legal texts sourced locally. NSTDA reports it performs on par with international competitors on Thai-context comprehension tasks. For businesses whose customers are primarily Thai nationals, that is a meaningful difference in quality when it comes to chatbots, content generation, document processing, and search. Chiang Mai businesses already investing in translation and localisation services will recognise this problem immediately. A model that genuinely understands Thai is a different tool from one that merely processes it.

What This Changes for Chiang Mai Businesses

The practical implications fall into several categories, and they are worth thinking through regardless of whether your business currently uses AI tools at all.

Thai-language customer communication. If you run a restaurant, guesthouse, retail shop, or service business with Thai-speaking clientele, an AI assistant built on ThaiLLM will handle enquiries, bookings, and FAQs with significantly higher accuracy than a standard ChatGPT integration. Local developers can now build these tools on a model trained on the same language your customers actually use.

Lower barriers to AI adoption for Thai-owned SMEs. ThaiLLM is free to access and free to download for customisation. Thai entrepreneurs who previously could not justify the cost or complexity of building AI-powered tools now have a foundation model they can adapt without licensing fees. Expect a wave of affordable, locally-built AI products aimed at the Thai SME market.

Data sovereignty and compliance. ThaiLLM is designed to run on local infrastructure, including the government’s ThaiSC supercomputing system. For businesses that handle sensitive customer data, particularly in finance, healthcare, legal services, and education, this matters. Processing data locally under Thai jurisdiction reduces regulatory exposure and aligns with emerging data localisation requirements.

Agentic AI on the horizon. NSTDA has already signalled that the next development phase involves agentic architectures, where specialised AI units are managed by a central controller to complete complex, multi-step tasks. This is the direction enterprise AI is moving globally. Thailand is building toward it locally, which means Thai-language business automation tools are coming as domestic products, not imports.

A new talent pool. Over 700 students and researchers have already completed AI training programs tied to the ThaiLLM initiative. That is a growing cohort of Thai developers with hands-on experience in foundation models. For Chiang Mai businesses looking to hire tech talent or partner with local developers, this pool is expanding.

The Chiang Mai Angle

Chiang Mai’s business community sits at an interesting intersection. The city has a substantial expat and digital nomad population comfortable with global AI tools, alongside a large Thai-speaking local economy that has historically been underserved by those same tools. ThaiLLM shifts that balance. That intersection is part of what makes Chiang Mai one of Southeast Asia’s most compelling business environments, and it is also what makes this development worth watching closely.

Major Thai financial and technology organisations, including KBTG and SCB 10X, have already begun integrating ThaiLLM into their own projects. When the country’s largest institutions move in a direction, the ecosystem around them follows. Local service providers, software developers, and agencies in Chiang Mai who want to build products for Thai enterprise clients should start paying attention to what ThaiLLM can do, and what it cannot yet do.

For businesses primarily serving the international market, the immediate impact is smaller. But the medium-term signal is clear: Thailand is building AI infrastructure at the national level. That changes the competitive landscape for local tech services, for marketing agencies selling AI-assisted solutions, and for anyone advising Thai business owners on digital tools. Thailand’s digital infrastructure is already a genuine competitive advantage — the country’s world-leading internet costs mean that cloud-based AI tools are accessible here in a way they simply are not in many other markets.

What Chiang Mai Businesses Can Do Now

1. Visit thaillm.or.th and run a Playground test. Try the queries your customers actually send you in Thai. Compare the output to ChatGPT or Claude. The differences will be instructive.
2. If you work with a local developer or digital agency, ask them whether they are building on or evaluating ThaiLLM. If they are not aware of it, that tells you something.
3. For businesses with Thai-language customer touchpoints, consider whether a ThaiLLM-based integration would outperform your current setup for customer service, booking systems, or FAQ tools.
4. If you are in professional services, particularly accounting, legal, or HR, track how ThaiLLM develops for document processing. Locally-trained models handling Thai contracts and compliance documents will be meaningfully more accurate than global alternatives.
5. Watch the NSTDA and BDI announcements. The agentic AI phase has been signalled. When it ships, the practical applications for business automation will expand quickly. For a broader view of how data and technology are already reshaping business strategy in the region, see our earlier piece on data insights driving tourism strategy in Chiang Mai.

A Note on What This Is Not

ThaiLLM is a foundation model. It is not a finished product and not a plug-and-play replacement for the AI tools you already use. Like all foundation models, it requires fine-tuning, integration work, and a clear use case to deliver value. The Playground is a demonstration, not a business tool in its current form.

The significance is structural, not immediate. Thailand now has a sovereign AI base layer. What gets built on top of it over the next two to three years is the story that will actually affect your business. The launch is the starting gun, not the finish line.

But the businesses in Chiang Mai that understand what has been built, who is building on it, and where the gaps still are will be better positioned to use those products when they arrive. Awareness is where it starts.

Sources and Further Reading

• Thailand Launches “ThaiLLM”: A Sovereign AI Foundation for the Nation — National Science and Technology Development Agency (NSTDA), April 2026
• ThaiLLM Playground and API — Big Data Institute / MDES
• Translation and Localisation Services in Chiang Mai — Chiang Mai Business Network
• Data Insights Reshaping Tourism Strategy in Chiang Mai — Chiang Mai Business Network
• Thailand Has the World’s Cheapest Internet in 2025 — Chiang Mai Business Network
• Chiang Mai Tops Safest Cities in Southeast Asia 2025 — Chiang Mai Business Network

CMBN Staff

chiangmaibusiness.net/

Amity, the AI and software company founded by Korawad Chearavanont in 2012, has just closed a $100 million Series D round — the largest generative AI capital raise in the region to date — bringing its total funding to $160 million. The round was led by EDBI, the investment arm of Singapore’s SG Growth Capital, alongside Asia Partners and SMDV. An IPO is targeted for mid-2027, with the Stock Exchange of Thailand, Singapore, and Hong Kong all under consideration.

The numbers are impressive. But the real story for Chiang Mai businesses is buried in the details of what Amity actually does, who its clients are, and where AI investment is flowing next.

Who is Amity and why does it matter here?

Amity isn’t building the kind of AI that makes headlines at Silicon Valley conferences. It builds what its founder calls vertical AI — artificial intelligence tailored to the specific operational needs of individual industries, primarily retail and telecoms. Its core products include enterprise chatbots, voice AI platforms, and agentic AI systems capable of running end-to-end business processes autonomously.

Its client list in Thailand reads like a map of the country’s dominant commercial infrastructure: CP All (7-Eleven), True Corp, Kasikorn Bank, AirAsia, and major government agencies. Between 35% and 40% of Amity’s revenue comes from Thailand. The rest comes primarily from Europe, where over 75% of its EBITDA is now generated — a remarkable feat for a company headquartered in Bangkok.

Korawad himself is the grandson of CP Group patriarch Dhanin Chearavanont, whose family Forbes ranks as Thailand’s second-wealthiest with a net worth of $29 billion. But he has been deliberate about building Amity independently, keeping family investment below 10% and seeking institutional scrutiny from external investors. “If the family just leads or funds,” he has said, “it will not have the kind of scrutiny that external institutional investors give.”

The CP Group connection and what it means for Chiang Mai

Here is where this story connects directly to the daily business environment in Chiang Mai.

CP Group’s commercial footprint in this city is enormous. Lotus’s supermarkets, Makro cash-and-carry, 7-Eleven convenience stores, True telecommunications — these are not background players. They are the primary suppliers, landlords, logistics partners, and service providers for thousands of Chiang Mai SMEs. A significant portion of the city’s retail supply chain runs through CP infrastructure in some form.

As Amity’s AI tools — customer service automation, voice bots, agentic business processes — are adopted deeper into CP Group’s operating companies, the knock-on effects will reach the businesses that supply, franchise, and partner with those companies. AI-powered procurement systems, automated supplier communication, AI-driven inventory management: these are not abstract future scenarios. They are already being piloted in Thailand’s largest retail networks, and the standards and processes they create will eventually become the baseline that suppliers are expected to meet.

For Chiang Mai businesses in the food production, logistics, retail, and hospitality sectors, understanding where Thailand’s dominant corporate infrastructure is heading technologically is not optional background reading. It is competitive intelligence.

The vertical AI signal — and what it means for your sector

Amity’s $100 million bet on vertical AI — AI built for specific industries rather than general-purpose tools — deserves attention from any Chiang Mai business owner thinking about their own digital transformation.

The era of “use ChatGPT for everything” is giving way to something more sophisticated and more useful: AI systems that understand the specific language, workflows, regulations, and customer behaviour of a particular industry. A chatbot built for a Thai telecoms company understands bill disputes, plan upgrades, and network outage protocols. One built for a hotel chain understands late check-outs, room type requests, and loyalty programme rules. These are not the same tool.

For Chiang Mai’s dominant industries — hospitality, food and beverage, wellness, artisan manufacturing, agricultural processing — the implication is clear. The most valuable AI investments in the next few years will not be generic. They will be tools that understand your industry’s context. The smart move now is to start defining what that context looks like for your business — what are your most repetitive customer interactions, your most time-consuming internal processes, your most common supplier communications — before the tools designed to automate them arrive.

A Thai success story as proof of concept

There is one more reason this story matters for Chiang Mai businesses, and it is perhaps the most important one.

Amity is a Thai company, built by a Thai entrepreneur, competing and winning in European markets with technology developed in Bangkok. Its revenue grew more than tenfold between 2022 and 2025, surpassing $100 million annually. It is now targeting a public listing on one of Asia’s major stock exchanges.

This matters because one of the persistent mental barriers to AI adoption among Thai SMEs — and Chiang Mai businesses are no exception — is the sense that this technology belongs to other people, other countries, other industries. That AI is something that happens to Thai businesses rather than something Thai businesses build and use on their own terms.

Amity is a direct rebuttal to that narrative. So, for that matter, is the depa OTOD AI Transformation programme currently offering up to 200,000 baht in grants to Chiang Mai SMEs willing to adopt digital tools before the 30 April 2026 deadline.

The technology is Thai. The funding is here. The tools are available. The question, as always, is whether local businesses will move before the window closes.

Related reading: Thailand Launches National AI Push — Here’s What Businesses Need to Know | Amity official website | Bangkok Post: Amity targets IPO after $100m AI funding round

CMBN Staff

chiangmaibusiness.net/

It’s called Ask Maps, and for any Chiang Mai business that relies on foot traffic, tourism, or local discovery — which is most of them — it matters more than almost anything else happening in digital right now.

What Ask Maps actually does

Until now, Google Maps worked the way it always has: you type a keyword, you get a list of pins. The algorithm decides which businesses surface based on proximity, relevance, and profile completeness.

Ask Maps changes the interface entirely. Instead of typing “coffee shop Nimman,” a visitor to Chiang Mai can now ask: “I want a quiet place to work with good espresso and fast wifi, not too crowded on a Tuesday morning.” Instead of searching “Thai massage near me,” they can ask: “Where can I get a traditional northern Thai massage that’s calm, not touristy, and has availability this afternoon?”

Gemini, Google’s AI, processes that question against its database of over 300 million places, synthesises reviews, photos, attributes, and operational details, and returns a curated, conversational recommendation — with a custom map, directions, and options to book or call directly. It also personalises results based on a user’s search and save history in Maps.

This is no longer keyword search. It is AI-mediated discovery. And the businesses it recommends will not necessarily be the ones with the most reviews or the closest location. They will be the ones whose information Gemini can understand, trust, and accurately describe.

The Chiang Mai context

Consider the questions that tourists, expats, digital nomads, and visiting buyers actually ask when they arrive in this city.

“Where can I find handmade silver jewellery made by local artisans, not mass-produced tourist stuff?”

“Is there a cooking school in the old city that does small groups and focuses on Northern Thai food, not just pad thai?”

“I need a co-working space near Nimman that has private meeting rooms and reliable enough internet for video calls.”

“What’s a good restaurant for a business lunch that’s quiet enough to talk, has English menus, and isn’t a tourist trap?”

These are real questions that real visitors ask. Until now, they asked them on TripAdvisor, in Facebook groups, and on travel forums. Ask Maps is now the place they will ask them — and Gemini will answer based entirely on what it can find in your Google Business Profile, your website, and the reviews your customers have left.

If that information is thin, incomplete, or absent — Gemini cannot recommend you. Not because it has chosen not to, but because it genuinely does not know enough about you to answer the question confidently.

What Ask Maps reads to make its decisions

Understanding what feeds Ask Maps is the key to acting on it.

Google’s AI draws primarily from three sources (right now) when generating Ask Maps responses: your Google Business Profile, your website content, and the body of customer reviews associated with your listing. These are not equally weighted. Your Google Business Profile is effectively the primary structured data source — the place where Gemini looks first for factual, reliable information about what you are, what you offer, where you are, and when you’re open.

Critically, AI now reads the text of reviews, not just the star ratings. If customers mention “quiet atmosphere,” “great for solo travellers,” “traditional Northern Thai,” or “good for business meetings” in their reviews, Gemini can use those specific phrases to match your business to the kind of conversational query described above. A business with 200 generic five-star reviews is less useful to Ask Maps than a business with 50 reviews that specifically describe what makes it distinctive.

Profile freshness also matters. Google’s own data shows that businesses which haven’t posted an update or added a photo in over 30 days experience measurable drops in AI visibility. In the Ask Maps era, your Google Business Profile is not a directory listing you set up once and forget. It is a live marketing channel that requires the same regular attention as your social media.

What you should do this week

The practical action list is straightforward, though it requires honest self-assessment.

Audit your Google Business Profile.

Open it now and ask yourself: does it accurately describe everything distinctive about your business? Are your categories correct and specific — not just “restaurant” but “Northern Thai restaurant” or “Khao Soi restaurant”? Are your services listed in detail? Is your description written in natural language that explains what makes you different, not just what category you fall into? Is your website linked and up to date? Are your photos recent, high quality, and genuinely representative?

Read your reviews as Gemini reads them.

Go through your recent reviews and note the specific words and phrases customers use. These are the signals Gemini is already using. Are they the signals you want? If your reviews mostly say “great food, will come back” without specific detail, consider how you prompt customers for more descriptive feedback — mentioning the specific dish, the atmosphere, the occasion.

Strengthen your website content

Your website is the second major data source Ask Maps draws from. A website that consists of a homepage with beautiful photos and minimal text gives Gemini very little to work with. Pages that explain your services in plain language, describe your location with specific neighbourhood references, answer common customer questions, and use the natural language your customers use when searching — these feed the AI with the structured, trustworthy information it needs to recommend you confidently.

Check your consistency across the web

Gemini cross-references multiple sources — your Google Business Profile, your website, TripAdvisor, Facebook Business, and other directory listings — to build its understanding of your business. Inconsistencies in your business name, address, phone number, or description across these platforms create uncertainty for the AI and reduce its confidence in recommending you. Consistency is a trust signal.

The deeper shift

Ask Maps is currently rolling out in the US and India, with global expansion planned over coming months. Thailand is not yet in the first wave. But the infrastructure for how Google’s AI understands and ranks your business is being built right now — and the businesses that have strong, complete, accurate, and regularly updated profiles when Ask Maps reaches Thailand will start with a significant advantage over those that scramble to catch up.

The broader pattern is one CMBN has been tracking for months. AI translation is dissolving language barriers for your international customers. The Thai government’s AI funding programme is putting digital tools within financial reach of even the smallest operators.

Thai-founded companies like Amity are demonstrating that AI is not a foreign technology but a Thai competitive advantage. And now Google Maps — the app that more of your potential customers use to find you than almost any other tool on the planet — has become an AI system that either knows your business well enough to recommend it, or doesn’t.

There is a version of this shift where Chiang Mai businesses benefit enormously. The city has an extraordinary concentration of distinctive, authentic, high-quality experiences that AI discovery tools are genuinely well-suited to surface — if the underlying information is there to work with.

That part is up to you.

Your Ask Maps checklist

• Claim and fully complete your Google Business Profile — every field, every category, every service
• Add fresh photos regularly — minimum once a month
• Write a business description in natural language that explains what makes you distinctive
• List specific services with descriptions, not just category names
• Respond to all reviews, especially those with descriptive detail
• Prompt customers to leave specific, detailed reviews mentioning what they experienced
• Ensure your website has clear, descriptive content about your services and location
• Check your business name, address, and phone number are identical across Google, Facebook, TripAdvisor, and other directories
• Search for your business on Ask Maps when it reaches Thailand — and treat the result as your report card

Related reading:Thailand Launches National AI Push — Here’s What Businesses Need to Know|Google’s Ask Maps announcement

CMBN Staff

chiangmaibusiness.net/

Three state agencies — the Digital Economy Promotion Agency (depa), the Revenue Department, and the Board of Investment (BoI) — have jointly launched the One Tambon, One Digital (OTOD) AI Transformation initiative, targeting 15,600 businesses and individuals by 2026 and aiming to generate an economic impact of at least 500 million baht.

The announcement marks a significant escalation in Thailand’s national digital strategy, moving beyond awareness campaigns into direct financial support for businesses at every level of the economy.

Why now?

Officials have been candid about the urgency. depa chief Nuttapon Nimmanphatcharin said the initiative is aimed at helping Thailand deal with fierce competition, labour shortages, volatile export conditions, and rapid changes in technology and consumer behaviour. depa has set a target of making Thailand’s businesses AI-ready by 2027, and the OTOD programme is the most concrete vehicle yet for reaching that goal outside Bangkok.

Two funding tracks — which one fits your business?

The scheme is divided into two practical streams, both accessible via aitransform.depa.or.th.

The first is for established SMEs ready to commit to transformation. The d-transform fund offers up to 200,000 baht per business for 600 qualifying SMEs — covering 200 manufacturing businesses, 300 trade and service businesses, and 100 agricultural processing operations. Grant rounds are reviewed on the 7th and between the 18th–22nd of each month, on a rolling basis until funds run out, and are capped at 250 grants divided equally across regions.

The second track targets smaller operators and first-time digital users. The d-voucher fund provides six months of initial digital usage rights for 15,000 participants — including small shops, street vendors, market stalls, and 5,000 farmers. Over 400 Thai-developed digital products and services registered on the Thailand Digital Catalog are available to choose from under this scheme.

The deadline has been extended — but time is running out

Applications for the AI Transformation programme are open now through 30 April 2026 at aitransform.depa.or.th. For telephone enquiries, depa can be reached at 061 225 7778 and 061 298 7774. Both funds operate on a first-come, first-served basis — don’t wait.

A new tax incentive adds further firepower

Separate from the OTOD grants but designed to work alongside them, Royal Decree No. 802, published in February 2026, introduces a 200% corporate income tax deduction for SMEs investing in digital products and services registered on depa’s Thailand Digital Catalog. The measure covers expenses incurred from 24 June 2025 through 31 December 2027.

In plain terms: spend 100,000 baht on a qualifying AI tool, and you can deduct 200,000 baht from your taxable income. Businesses need to retain official tax invoices, proof of payment, and depa registration confirmation for each qualifying expense — so keeping clean records from the outset is essential.

Chiang Mai is one of seven roadshow cities

Workshops, seminars, and business-matching events are being organised across seven provinces: Phitsanulok, Chiang Mai, Songkhla, Phuket, Chon Buri, Ubon Ratchathani, and Khon Kaen. These events connect local businesses with vetted digital solution providers from the Thailand Digital Catalog — valuable for operators who aren’t sure which tools to choose. The national roadshow has now reached its eighth stop, confirming the programme is actively rolling out across the country.

depa has regional offices in Chiang Mai where staff can assist with grant applications, and the agency has explicitly stated that retail, agriculture, tourism, and food businesses are among its priority targets — not just technology companies.

Who qualifies?

SMEs with annual revenues under 600 million baht or fewer than 300 employees are eligible for the transformation grants. For the d-voucher, the bar is lower — it is designed for micro-businesses, sole traders, and farmers accessing digital tools for the first time. The agency anticipates over 15,600 entrepreneurs across Thailand will benefit from the scheme.

The larger context

Research into Northern Thailand SMEs specifically identifies financial constraints, workforce skill shortages, and poor data quality as the three biggest barriers to AI adoption — precisely the gaps the OTOD programme is designed to close. Thailand’s digital transformation market is projected to nearly double from US$9.2 billion in 2024 to nearly US$20 billion by 2033, with Chiang Mai identified as one of the country’s key regional growth hubs.

For local businesses, the message from Bangkok is clear: the window is open, the funding is live, and the deadline is 30 April 2026.

How to apply

• Programme portal and d-voucher applications: aitransform.depa.or.th
• depa official website: depa.or.th
• Follow updates: depa Thailand on Facebook
• Phone enquiries: 061 225 7778 / 061 298 7774
• Background reading: Bangkok Post original report | Royal Decree No. 802 explained | Northern Thailand SME AI research

Sources: Bangkok Post, depa Thailand, Royal Gazette — Royal Decree No. 802, ScienceDirect Northern Thailand SME Study, Pertama Partners Digital Economy Guide

CMBN Staff

chiangmaibusiness.net/

The answers matter for anyone running a business in Chiang Mai, because they reveal not what AI vendors are selling, but what real people are trying to solve. And the picture that emerges is more grounded, more human, and more immediately useful than most of the headlines about AI would suggest.

Here is what the data shows, and what it means if you are operating an SME in northern Thailand.

The top thing people want: time back

According to the research, time-saving was the most cited benefit, mentioned by half of all respondents. Not intelligence. Not automation in the abstract. Time.

But the follow-through is telling. When Anthropic asked people what they would do with that recovered time, the answers were not about more work. People described picking up children from school, cooking with family, reading books, and being present. The full findings make clear that productivity was rarely the end goal. It was the means to a better life outside work.

For Chiang Mai business owners juggling operations, customer service, admin, and marketing, this is the honest case for AI adoption. Not the transformation pitch. The practical one: what are you doing right now that a tool could handle, so that you get an hour back today?

The second most wanted thing: help doing better work

“Professional excellence” was the largest single category of AI vision, cited by nearly one in five respondents. People wanted AI to handle routine tasks so they could focus on higher-value strategic work and complex problem-solving.

This maps directly onto what most Chiang Mai SME owners already feel. The work that matters, the relationships, the judgment calls, the creative decisions, tends to get crowded out by the work that just has to get done. Drafting the same reply for the tenth time. Compiling a report no one reads. Translating a document. Formatting a quotation.

AI does not replace the parts of your job that require you. It creates space for them.

Entrepreneurs are the most optimistic group in the study

Among people who reported real economic gains from AI, independent workers and entrepreneurs stood out sharply. Nearly half reported concrete economic empowerment, compared to 14% of institutional employees.

People in lower and middle income countries also showed significantly more optimism than those in Europe or North America, framing AI as a capital bypass mechanism: a way to start and grow businesses without the funding, hiring infrastructure, or specialist access that would otherwise be required.

That framing is directly relevant here. Running a business in Chiang Mai often means operating without the specialist depth available to larger companies. No in-house marketing team. No dedicated accountant on staff. No full-time content writer. AI does not fix all of that, but it compresses the gap.

If you are still weighing up the fundamentals, our guide to starting a business in Chiang Mai covers what the city actually offers operators at this stage.

The biggest concern: unreliability

The most common worry in the study was AI unreliability, raised by more than one in four respondents. People were concerned about hallucinations, inaccurate outputs, and the verification burden that comes with having to fact-check everything an AI produces.

This is legitimate, and it is something to take seriously rather than dismiss. AI tools produce confident-sounding output that is sometimes wrong. For any business task where accuracy matters, particularly anything involving numbers, legal compliance, or customer-facing claims, human review is not optional.

The practical implication is not to avoid AI, but to use it in the right parts of your workflow. AI is strong on drafting, structuring, translating, summarising, and generating options. It is not a substitute for judgment, verification, or domain expertise.

Translation and bilingual content is one area where the time saving is immediate. If you need professional support alongside it, see our translation services directory.

The second biggest concern: jobs

Concern about job displacement and economic disruption was raised by more than one in five respondents, and it was the single strongest predictor of negative overall AI sentiment in the study.

For business owners, this concern plays out in two directions. There is the question of what AI means for your own team, and the question of what it means for your own competitive position.

On the team side, the honest answer is that AI will change some job functions significantly. The businesses that handle this well will be the ones that help their staff use AI to do better work. Getting your team structure right first matters — our piece on employee turnover costs is worth a read before making any staffing decisions.

On the competitive side, freelancers and small business owners were the most exposed group in the study, simultaneously benefiting from AI and feeling precarious because of it. They are both users of the tools and, in some sectors, competing against them. That tension is real and worth taking seriously.

A pattern that should inform how you adopt AI

One of the more useful findings in the study is about where AI’s benefits are strongest.

Tradespeople reported some of the highest rates of learning benefits from AI, yet almost none reported signs of cognitive atrophy. The researchers concluded that AI’s benefits may be strongest when learning is volitional, as opposed to within institutional structures where AI is more likely to be used as a shortcut.

Applied to a business context: AI tools produce the best results when the person using them already has judgment in the area. A business owner who understands their customers will use AI to write better marketing copy. Someone with no feel for their audience will produce polished nonsense faster.

This is an argument for building AI into your workflow as a complement to your expertise, not as a replacement for developing it.

What Southeast Asian users specifically want

In Southeast Asia, respondents showed lower concern levels around governance, surveillance, and abstract AI risk compared to Western regions, indexing more strongly toward practical concerns about reliability and economic impact, and toward a vision of AI as a tool for creating opportunity rather than managing complexity.

That framing, AI as opportunity engine rather than productivity optimiser, is a useful one for Chiang Mai businesses thinking about where to start. The question is not how to plug AI into an existing workflow. It is where AI opens up something that was not previously accessible: a new market, a new capability, a new customer segment you could not serve before.

Where to start if you have not already

The Anthropic data points toward a few high-value entry points for SMEs.

Writing and communication is the most immediate. Drafting emails, social media posts, customer replies, and internal documents are areas where AI can produce a strong first draft that you edit and approve. The time saving is real and the verification burden is low.

Research and synthesis is the second. If you regularly need to understand something new, a regulation, a market, a competitor, or a technical process, AI is a fast first-pass research tool. Treat its output as a starting point, not a final answer.

Translation and bilingual content is a significant opportunity in Chiang Mai specifically, given the daily need to communicate across Thai and English. AI translation is not perfect, but for internal communications and first drafts it dramatically reduces the time cost of operating bilingually.

Operational templating is the fourth. Standard operating procedures, onboarding documents, quotation templates, and routine correspondence can all be drafted with AI assistance and then refined once rather than recreated repeatedly.

For a broader look at how to position your business online in 2026, see our website marketing strategies guide.

The summary

The Anthropic study did not find a world divided into AI believers and sceptics. It found people organised around what they value, watching AI develop while managing both hope and anxiety at once. The most common pattern was not optimism or pessimism, but a set of genuine tensions: the same tools that save time can also increase pace; the same tools that empower entrepreneurs can displace workers in adjacent roles.

For Chiang Mai business owners, the most productive way to engage with that complexity is to start with a specific problem rather than a general position on AI. What is the task in your business that takes the most time relative to the value it produces? That is the place to test a tool. Measure the result. Go from there.

The businesses that will gain the most from AI in the next two years are not the ones that adopted it earliest. They are the ones that adopted it most deliberately.

CMBN Staff

chiangmaibusiness.net/

Then ChatGPT arrived…..

A Brief History of the Mark

The em dash has been part of written English for centuries. It takes its name from its width — equal to the letter ‘m’ in typeset text — distinguishing it from the shorter en dash (the width of ‘n’) and the humble hyphen. Early typesetters used it to indicate pauses, interruptions, and abrupt shifts in thought. It appears in the work of Emily Dickinson, who used it so frequently and idiosyncratically that it became a defining feature of her poetry — nearly 1,800 dashes across her body of work, each one doing something a full stop or comma could not.

Jane Austen used it for narrative aside and irony. Charles Dickens used it for dramatic interruption. Henry James, never accused of brevity, used it to hold subordinate thoughts inside larger ones without breaking the sentence apart. In American literature, the em dash found particular favour — looser, more conversational, more willing to let a sentence breathe mid-way through.

In formal British writing, it has always sat a little awkwardly. More informal than the colon, more emphatic than the comma, it occupied a space between the two — useful precisely because it did not quite belong to either.

Who Used It Well

The writers most associated with the em dash used it with purpose, not decoration.

Emily Dickinson made it the structural backbone of her verse. Where others used punctuation to clarify, she used it to create suspension — the reader held in mid-air between one thought and the next.

Joan Didion, one of the most precise stylists in American journalism, used it to place the personal inside the political — an aside that reframed the whole sentence. Her dashes were never accidental.

David Foster Wallace used it to replicate the fractures of actual thought — the half-finished idea, the sudden qualification, the moment where language catches up with thinking. In his hands it was almost conversational.

In business and journalism, good writers used it to separate, to emphasise, to redirect. It was the mark that said: pay attention here, this part matters.

What AI Did to It

The em dash did not change. What changed was who started using it, and how.

Large language models — ChatGPT in particular — use the em dash relentlessly. It appears in AI-generated text at a frequency that no human writer would sustain. Open any AI-written article, marketing copy, or business document and you will find it. Multiple times per paragraph. Sometimes multiple times per sentence. Not used with the deliberateness, but scattered throughout like a tic — a signal, it turns out, of automated generation rather than human craft.

The result is that a mark which once indicated stylistic intention now indicates the opposite. Readers have learned — quickly — to associate the em dash with content that was not written by a person. Whether that association is entirely fair is beside the point. Perception is reality in publishing, and the perception is clear.

If you use an em dash now, a significant portion of your readers will assume you did not write the sentence yourself.

The Problem for Those of Us Who Used It First

This is the frustration. I was using the em dash before most of my readers had heard of ChatGPT. It was part of my voice — earned, not generated. And now I must retire it, not because it stopped working, but because someone else’s tool adopted it so aggressively that it no longer reads as human.

It is the linguistic equivalent of a phrase becoming so overused it loses meaning. Except this time, it happened in under two years, and the culprit was a machine.

How to Stop Using It

If you are in the same position, the transition requires more thought than simply finding and replacing every dash. The em dash did specific work. Each use needs a different solution.

Where you used it to add emphasis or a dramatic aside, try a colon. It is more formal but it carries authority: it promises that what follows matters. Where you used it for a parenthetical thought, use actual parentheses, or restructure the sentence so the aside becomes its own clause. Where you used it to show an abrupt change in direction, try a full stop and start a new sentence. The pause becomes more emphatic, not less. Where you used it to set off a list embedded in a sentence, a comma usually works. Occasionally restructuring around a semicolon gives you the separation you were after.

The deeper adjustment is stylistic. The em dash allowed a certain looseness — a willingness to let thought interrupt itself mid-sentence. Without it, writing becomes more composed, more deliberate. Sentences either hold their structure, or they break into separate ones. There is less improvisation on the page. That is not necessarily a loss. It is a discipline.

The Honest Conclusion

The em dash is not gone. It still exists in style guides. It still has its place in fiction, in informal writing, in contexts where a human voice is unambiguous. But in professional writing, in articles, in business communications, in anything where your credibility as the author matters — it has been compromised.

Retire it for now. Not forever. But until the association fades, which may take years, every dash you use is a question mark over your authorship. And that is too high a price for a punctuation mark, however useful it once was.

Use your words. Not AI’s favourite decoration.

Lone Andersen

Business Advisor | Champion of Strategic Growth & Sustainable Innovation

Lone Andersen is a dynamic business leader, serial investor, and startup founder with a global track record of driving growth and sustainability. From advising governments on waste management in Singapore, Rwanda, and Bangladesh to scaling B2B and B2C ventures across Asia, Europe, and Australia, Lone’s expertise spans industries and borders. Known for her sharp strategic insight, she empowers founders, investors, and startups to establish and expand in Thailand and ASEAN. With a passion for sustainable business practices, Lone is the trusted partner for those aiming to scale smart, grow sustainably, and lead with impact.

chiangmaibusiness.net