AI Everything Abu Dhabi 2026 runs 6 to 7 October at ADNEC Centre. I counted the official exhibitor directory, picked the stages I would sit in, and wrote down how I would plan the two days.
AI Everything Abu Dhabi 2026 runs on 6 and 7 October at ADNEC Centre, Halls 1 to 4, hosted by Abu Dhabi's Department of Government Enablement. Before writing this, I counted the official exhibitor directory. It lists 298 companies and public bodies. 161 of them sit at startup pods, and 165 are in a single hall.
I have not walked this floor. Everything below comes from the organiser's published material, the directory, two talks on YouTube, and what I have learned working rooms like KubeCon India, WeAreDevelopers in Berlin, and HumanX Amsterdam. Read it as a planning letter: what the event is, where I would spend the hours, and what I would do the week after.
If you are exhibiting, one more question is waiting for you after the show. When a visitor flies home and asks ChatGPT or Perplexity about your category, does your company come up? We built a free AI visibility index for AI Everything Abu Dhabi 2026 so you can check.
| Quick fact | Detail |
|---|---|
| Dates | Tuesday 6 and Wednesday 7 October 2026 |
| Hours | Tuesday 8:30 AM to 6 PM, Wednesday 9 AM to 5 PM (plan your visit) |
| Venue | ADNEC Centre Abu Dhabi, Halls 1 to 4 |
| Host and partner | Department of Government Enablement (host), Solutions+ (global partner) |
| Day one stages | AI Summit: New Intelligence, plus Intelligent Infrastructure & New Energy |
| Day two stages | AI Government Excellence Forum, plus AI 360 |
| Hands-on, both days | AI workshops, certified trainings, AWS AI Training Academy, Alibaba Qoder AI Hackathon |
| Startups | AI StartX zone, with the Supernova Challenge final on 7 October |
| Getting there | Etihad Rail from Dubai in 57 minutes, from AED 39. Careem code AIE2026C and Uber code AIE2026U give 15% off two rides |
| Event app | AI Everything+ |
TL;DR
- AI Everything Abu Dhabi 2026 runs 6 to 7 October at ADNEC Centre, Halls 1 to 4, hosted by Abu Dhabi's Department of Government Enablement: a government-hosted AI expo built for people who buy, fund or regulate AI in the Gulf.
- The official exhibitor directory lists 298 companies, well below the 850+ once promised in a 2025 promo. 161 of those are startup pods, and 165 sit in Hall 4 alone, so plan around the directory, not the headline number.
- Compute cost is the recurring theme: Core42's Talal Al Kaissi puts one gigawatt of AI data centre capacity at roughly $45 billion to build, and the UAE has already committed to 5 gigawatts.
- A second thread runs underneath it: speakers from Mistral AI, Multiverse Computing, and ModelBest will make the case for smaller, cheaper models over frontier-scale ones.
- For a founder at a pod, the real value is buyer language collected before and during the show, turned into content within 48 hours, which is what Infrasity's free AI visibility index for this event is built to check.
What is AI Everything Abu Dhabi 2026, and who is it built for?
It is a government-hosted AI expo and summit, built for people who buy, fund or regulate AI in the Gulf. The host is the department leading Abu Dhabi's move to an AI-native government. The organiser's own audience list starts with government and policymakers, then C-suite and enterprise buyers, then investors.
That order shapes what gets said on stage. The opening-day summit, branded New Intelligence, covers sovereign AI, agentic systems and AI-native economies. Sovereign AI means a country keeping its data, models and compute under its own law. The speaker list includes the UAE's Minister of State for AI, Omar Sultan Al Olama, Arthur Mensch of Mistral AI, Talal Al Kaissi of Core42, and Boston Dynamics founder Marc Raibert.
Here is how I would place it next to the events we wrote about this year:
- The AI Conference in San Francisco was an engineer's room: agent failures, evaluation, inference cost.
- HumanX Amsterdam was a room of VPs deciding on AI spend.
- Abu Dhabi sits one level above both. The buyer is often a ministry, and the budget is national.
One planning note. The dates moved. The 2026 edition was first announced for 11 to 13 May, and some event listings still show older dates. The ADNEC venue listing confirms 6 to 7 October in Halls 1 to 4.
What does the exhibitor directory show that the brochure does not?
The directory shows a floor you can cover in two days, and it shows where the startups are. I pulled every entry from the official exhibitor directory on 5 October 2026 and removed two test records. That left 298 listings.

The exhibitor directory breakdown: 55% of listings sit in Hall 4 alone. Source: official exhibitor directory, snapshot 5 October 2026.
What the count says
- 161 of the 298 listings are pods: 128 sit in Hall 4 and 33 in Hall 3.
- Hall 4 holds 165 listings: That is more than Halls 1, 2, and 3 combined.
- 150 listings are UAE-registered: India and the US follow with 36 each. Country means the registered entity, so Microsoft Gulf and Cloudflare Middle East count as UAE.
- 31 of India's 36 listings are pods: The Indian presence is almost all startups.
- 60 listings tag themselves "Chatbots / Virtual Assistant": That is one in five, and 41 of them are pods.
A note on method. The directory does not label pods. I am reading a P in the stand code as a startup pod because 18 of the 22 startups I could match from the organiser's AI StartX page sit at P-numbered stands.
Plan around the directory, and ignore the headline number
The organiser's numbers have moved over the year. A promo video from late 2025 promised 850+ AI companies. The homepage now says 350+. The directory lists 298. I would plan around the directory, because it is the only one of the three with stand numbers.

A stand code tells you the stand size before you even open the profile: a P means a startup pod, a letter plus number means a full stand.
A smaller floor is good news for a visitor. The show is open for 17.5 hours across the two days. Spread over 298 listings, that is about 3.5 minutes each, with no sessions and no lunch. So shortlist before you arrive. Use the directory's country and category filters, pick 15 stands, and write the stand codes down.
Which stages and speakers would I prioritise?
I would follow one thread across both days: what AI costs to run, and who is building the cheaper version. Two groups of speakers carry it.
The compute bill, counted in gigawatts
My first pick is Talal Al Kaissi of Core42. His interview at the 2025 summit is the most useful 17 minutes I found from this event, because he put numbers on the problem:
- About 60 gigawatts of data centre capacity existed worldwide when he spoke.
- Serving AI to the whole world would need 300 to 500 gigawatts, by G42's calculation.
- One gigawatt costs roughly $45 billion to build and fill with GPUs.
- A cloud rack drew around 17 kilowatts. AI racks now need more than 100, with liquid cooling.
Intel's Middle East lead, Taha Khalifa, adds the local figure in a pre-event interview: the UAE has committed to 5 gigawatts, and one gigawatt is close to the output of a nuclear plant.
In Berlin, the cost conversation was about tokens per task. Here, the same conversation is counted in gigawatts. The question I would bring to the Intelligent Infrastructure stage: what has changed in those numbers since February 2025?
For founders, the takeaway is practical. Your buyer in this region thinks about where the compute sits and whose law covers the data. Have an answer to both before they reach your stand.
The smaller, cheaper model
My second pick is a group. Arthur Mensch runs Mistral AI. Enrique Lizaso Olmos runs Multiverse Computing, which compresses large models. Li Dahai founded ModelBest, which builds small language models that run on phones and cars. All three are on the speaker list.
The organiser is already framing this debate. Ramanan Ramanathan of the IndiaAI Mission argues the next phase will be won by the most trusted model, ahead of the largest one. I would sit in these sessions with one question: at which task does the small model stop being good enough?
AI 360 on day two
AI 360 covers enterprise deployments that have moved past the pilot. I would apply the same test I suggested for The AI Conference: ask the speaker to walk through one user task, including the path where it fails.
Session times live on the official agenda, and the AI Everything+ app builds your schedule. One detail to plan around: the organiser lists Wednesday closing at 5 PM, while the venue listing says 6 PM. Plan for 5.
How should a founder at a pod use the two days?
Treat the pod as a place to collect buyer language, and book your meetings before you arrive. Most of the value at an event like this is decided before the doors open and in the week after.
Before the doors open
- Write down three buyer roles: A ministry digital lead, a bank's head of data and a telecom CIO need three different first sentences.
- Request meetings through the app: The Chief AI Officers Programme admits members by invitation or approved application, and each one commits to at least five confirmed meetings. Those calendars fill before the doors open, so do not count on a walk-up.
- Use the investor matchmaking: The Investors Programme runs on pre-scheduled, concierge-matched meetings. The organiser lists 200+ investors, including Shorooq, Antler, Plug and Play and Global Ventures.
On the floor
- Open with the buyer's problem: You are one of 161 pods, and 60 exhibitors describe themselves as chatbots or virtual assistants. A category label will not get you remembered.
- Ask one question in every conversation: how did you shortlist your last AI tool? Write the answer down word for word.
- Watch the Supernova final on 7 October, even if you are not in it. The format is five finalists, a 2-minute pitch, 3 minutes of questions, and one $10,000 equity-free prize. Judges score market opportunity, business model, and team. Two minutes is about what a pod visitor gives you, so borrow the structure.
The week after
- Follow up within 48 hours, while the stand code and the conversation are still fresh for both sides.
- Turn buyer questions into pages. Documentation, comparison pages, and technical posts, written in the words you collected.
- Check where buyers look next. The B2B buyer journey now often starts in an AI assistant, so test whether those assistants name you for the questions you heard.
Is your company showing up in AI answers after the show?
A badge scan gets you a name. What the visitor does next decides whether it becomes a conversation, and that next step is often a question typed into an AI assistant.
This is for everyone exhibiting at or visiting AI Everything Abu Dhabi. Our AI Everything Abu Dhabi AI Visibility Index tests buyer questions across ChatGPT, Claude, Perplexity, and Gemini, then shows how often each engine names your company. The questions are unbranded. We ask about the problem your product solves and never give the assistant your name.
Find your company to see its results across the four engines. You can also request the full report, which shows the tested question, the competitors named in your place, and the sources the engines cited. If you are exhibiting and your company is not listed, request a report and tell us the question your buyers ask. The benchmark is free.

Infrasity's AI visibility index ranks companies by how often ChatGPT, Claude, Perplexity, and Gemini name them for unbranded buyer questions.
Check your company's AI visibility
How Infrasity helps after the event
A low score usually points to a specific gap: a missing answer page, thin comparison content, or third-party sources that mention competitors in your place. Closing those gaps is the work we do for DevTool and AI companies through our AI GEO optimization service, with content written by engineers.
Measured the same way as the index, ScaleKit was cited across 878 AI answers on 225 buyer prompts, and Firefly went from 0 to 12 AI citations in twelve months. If you are in Abu Dhabi this week and want to talk through your result, book a call with us.
What would I want to take home from Abu Dhabi?
Three things:
- An updated price for a gigawatt, and a clear view of who is paying for it.
- Fifteen conversations with a stand code next to each, and the exact words every buyer used.
- An answer to one question: when a ministry or a bank in this region shortlists an AI vendor, where does it look first?
My bet is that "sovereign" will be the word heard most often in Halls 1 to 4. The events we covered earlier this year asked how to build with AI at a lower cost. This one asks who owns the infrastructure it runs on.
Frequently Asked Questions
When and where is AI Everything Abu Dhabi 2026?
It runs on Tuesday 6 and Wednesday 7 October 2026 at ADNEC Centre Abu Dhabi, in Halls 1 to 4. Doors open at 8:30 AM on Tuesday and 9 AM on Wednesday, according to the organiser's Plan Your Visit page.
Did the dates for AI Everything Abu Dhabi 2026 change?
Yes. The edition was first announced for 11 to 13 May 2026 and later moved to October. The ADNEC venue listing and the official site both show 6 to 7 October 2026. If a listing shows a different date, treat it as out of date.
How many exhibitors are at AI Everything Abu Dhabi 2026?
The organiser's homepage says 350+ enterprises and startups. The official exhibitor directory listed 298 companies and public bodies when I counted it on 5 October 2026, after removing two test entries. 161 of those are at startup pods.
What is the Supernova Challenge at AI Everything Abu Dhabi?
It is the event's startup pitch competition. Five finalists each give a 2-minute pitch followed by 3 minutes of questions, and one winner receives $10,000 in equity-free funding. The final is on 7 October, and only exhibiting startups can apply.
Which passes are available for AI Everything Abu Dhabi 2026?
The organiser's FAQ lists Visitor, Training, Workshop Class, All Access, and Premium passes. Passes are non-refundable and non-transferable, and they can be upgraded. Check the registration page for current prices.
Can I get an AI visibility report if my company is not in the index?
Yes. If your company is exhibiting, request a report from the index page. You can also suggest the question your buyers ask, so the test reflects your use case.
How can Infrasity help if my AI visibility score is low?
The full report names the competitors that appeared in the tested answers and the sources those answers cited. We use that as the starting point for a review of your technical content, comparison pages and buyer-question coverage. A low score reflects the questions tested, and a different question can give a different result.








