- Prepare evidence and lettersUsually the longest partWeeks to months
- Apply online to the Home OfficeEndorsement fee£561
- Endorsing body decidesDepends on the endorsing body2 to 8+ weeks
- Apply for the visa£205 visa fee plus healthcare surchargeWithin 3 months
| Point | Tech Nation (digital technology) | Peer review (Royal Society or Royal Academy of Engineering) |
|---|---|---|
| Typical fit | Builds or leads AI products and platforms | Active researcher in an academic, industry or government research institution |
| Entry condition | Recognition in digital technology in the last 5 years | Active researcher with a PhD or equivalent research experience |
| Papers | Can support the academic contribution criterion | Central to the assessment |
| Letters | 3 letters from established digital technology experts | 1 letter from an eminent person living in the UK, plus a second letter for Exceptional Talent |
Endorsing body
Three endorsing bodies list AI and machine learning
GOV.UK's list of technical skills covered by Tech Nation includes "Artificial Intelligence, Natural Language Processing and Machine Learning experts (AI, NLP, ML)". The GOV.UK disciplines list for science, engineering, humanities and medicine places artificial intelligence and machine learning under the Royal Society, and the Royal Academy of Engineering lists artificial intelligence, machine learning and data science among its disciplines. The same job title can therefore lead to different endorsing bodies. What decides it is the kind of work you do and how it has been recognised.
- Building, deploying or leading AI products at a product-led digital technology company points to Tech Nation.
- Research in an academic, industry or government research institution points to the Royal Society or the Royal Academy of Engineering.
- Engineering-focused research, for example robotics, autonomous systems or computer vision for industry, often sits with the Royal Academy of Engineering.
Research routes
Peer review: the route for active researchers
The peer review route asks you to be an active researcher, to hold a PhD or equivalent research experience (which can include industrial research), and to work in an approved research discipline. The Royal Society and the Royal Academy of Engineering assess your research record rather than Tech Nation's digital technology criteria.
- Peer review needs a letter of recommendation from an eminent person living in the UK who knows you and your work.
- Exceptional Talent applicants also need a second letter from a senior member of a UK organisation, who does not need to know you personally.
- These letter rules are different from Tech Nation's three letters, so decide the route before you approach anyone.
Tech Nation route
Tech Nation: for AI work that builds products
Tech Nation looks at digital technology work, usually in-house at a product-led digital technology company. Technical applicants must show proven technical expertise with the latest technologies in building, using, deploying or exploiting a technology stack and building technical infrastructure. You need the mandatory criterion, recognition by others as a leading talent or potential leading talent in digital technology in the last 5 years, plus 2 of 4 optional criteria.
- Academic contribution is one of the four optional criteria: research published or endorsed by an expert, such as a paper in a top-tier peer-reviewed journal or presentations at a peer-reviewed conference.
- An undergraduate or MSc thesis does not qualify for that criterion.
- For Exceptional Promise, the Rules say research endorsed by an expert, not merely published.
Reading the rules is one thing. Knowing how they apply to your own evidence is another.
Book a free 30-minute callChoosing
How to tell which route fits your record
Look at what others recognise you for. If your strongest evidence is papers, citations, peer-reviewed conference work and research grants, a research route is usually the natural fit. If it is models in production, products, open source libraries and talks at technology events, Tech Nation usually is. Many AI researchers in industry have some of both, and the route with the stronger evidence is the one to pick.
- Do not split one record across two routes: you apply to one endorsing body at a time.
- A researcher named on an endorsed funder grant or fellowship may also qualify through UKRI.
- If you are unsure, the free eligibility check names the endorsing body most likely to assess you.
Documents and letters
Evidence limits and the three letters
You can send up to 10 pieces of evidence. At least 2 must show recognition for the mandatory criterion and at least 4 must cover 2 optional criteria, 2 for each. Each piece can be up to 3 sides of A4, and you cannot use the same piece of evidence for more than one criterion. Your CV and letters do not count towards the 10.
- Three letters of recommendation from three different well-established experts in digital technology, with detailed knowledge of your work over a period of 12 months or more.
- Tech Nation says reference letters from an immediate colleague, manager or friend are not sufficient as evidence of recognition.
- Vague or generic letters are one of Tech Nation's common reasons for non-endorsement.
Common mistakes
Mistakes that weaken AI and ML applications
These come from Tech Nation's published guidance and the peer review rules.
- Applying to Tech Nation with a record that is mostly papers, or to peer review without being an active researcher.
- Evidence created only to support the timing of the application, such as a burst of papers, repositories or talks just before applying.
- Showing what the lab or company achieved without showing your personal work.
- Relying on LinkedIn or Medium posts, which Tech Nation says are not sufficient for the mandatory criterion.
- Counting a patent application: Tech Nation says patents used as evidence must be granted.
Related guides
Keep reading
Guides applicants usually read next on this topic.
Global Talent Visa UK: The Complete 2026 Guide
The Global Talent visa needs no job offer or sponsor and lets your family join you. Stage 1 is the endorsement application, where an endorsing body decides whether you are a leader or a potential leader in your field.
Read related guideGlobal Talent Visa UK Requirements 2026: Who Can Apply
Each endorsing body sets its own Stage 1 tests, so the requirements for a software founder, a researcher and a theatre director are very different.
Read related guideGlobal Talent Visa Documents Checklist for Stage 1
Each endorsing body has its own document rules, from page limits to who can write your letters.
Read related guideThis guide is part of our complete Global Talent visa guide, which covers who the visa is for, the endorsing bodies, costs, decision times and settlement.
Applying from outside the UK. This guide covers the Stage 1 endorsement for AI and machine learning work under UK rules as published on 17 November 2026.
Article FAQ
Questions to settle before you spend time or money.
Use these answers to check the practical risks, documents, and decisions connected to this topic.
Which endorsing body assesses AI researchers?
It depends on your work. Tech Nation lists AI, NLP and ML experts as technical skills. The Royal Society and the Royal Academy of Engineering list artificial intelligence and machine learning as research disciplines, but only for active researchers with a PhD or equivalent research experience.
Do I need a PhD?
For the research routes, you need a PhD or equivalent research experience, which can include industrial research. Tech Nation does not require a PhD; it looks at recognition in digital technology and the optional criteria.
Can my papers count for Tech Nation?
Yes, as one of the four optional criteria: academic contribution through research published or endorsed by an expert. You still need the mandatory criterion, a second optional criterion and three letters.
I work in an industry AI lab. Which route?
Either can fit. If your work is research and recognised through publications and peer-reviewed conferences, a research route may suit you. If it is building and deploying products, Tech Nation is more usual.
Who can help me choose the route?
Tier 1 Global Talent Visa, led by founder Mihir Thaker, works only on the Stage 1 endorsement. We compare your record against each endorsing body's published criteria and plan the evidence and letters. The first call is free and lasts 30 minutes.
Stage 1 help
Find out which endorsing body fits your AI record
We help only with the Stage 1 endorsement. We compare your research and product evidence against each route and plan the letters the right endorsing body asks for.
Which field is your work in?
