London Becomes AI's Second Capital, 1300+ Jobs Open as Anthropic and OpenAI Lock In Mega Leases

London Becomes AI's Second Capital, 1300+ Jobs Open as Anthropic and OpenAI Lock In Mega Leases

The two most valuable AI labs on the planet just doubled down on London. Anthropic has signed a 158,000 square foot lease at One Triton Square, with capacity for 800 people, while OpenAI committed to 88,500 square feet for more than 500 staff in King's Cross. That is roughly 1,300 new desks in a single AI cluster, announced inside two weeks of each other. For job seekers paying attention, this is the clearest signal yet that London has become AI's second capital after San Francisco, and the hiring window is open right now. If you are thinking about your next move, Metaintro is tracking this push closely, because the roles being posted today will shape the next decade of work.

Two Mega Leases, One Knowledge Quarter

Anthropic's new King's Cross headquarters is roughly four times the size of its current UK footprint, jumping from about 200 staff to a planned 800 by late 2026 or early 2027. The deal, signed with British Land and Royal London Asset Management, lands the Claude maker in the same block as Google DeepMind, Meta, Wayve, Synthesia and Isomorphic Labs. The cluster effect matters. AI engineers tend to job hop within walking distance, and a 158K square foot Anthropic flag pulls every senior researcher in the area into a recruiter's reach.

OpenAI's lease is smaller in square footage but symbolically heavier. After halting its UK Stargate infrastructure project earlier in the year, OpenAI signaled that London is now the company's largest research hub outside the United States. The new 500-plus-seat office is OpenAI's first permanent UK base, and it sits inside the same Knowledge Quarter cluster. According to CNBC, OpenAI currently employs around 200 people in London, so the lease implies more than doubling its UK headcount.

What Roles Are Actually Opening

The two companies are not hiring the same way. Anthropic is leaning into research, alignment and safety, with about 60 AI safety researchers already on the London team and a stated plan to deepen work with the UK AI Security Institute. Expect heavy intake on machine learning research engineers, interpretability researchers, reinforcement learning specialists, security engineers monitoring frontier models, and policy leads who can interface with UK and EU regulators. London will also be Anthropic's primary EMEA enterprise base, so go-to-market hiring for solutions architects, account executives covering UK financial services and German manufacturing, and customer engineers who can deploy Claude inside regulated industries is already active.

OpenAI's London push is tilting toward applied research and EMEA commercial. The company has been recruiting ChatGPT product engineers, post-training researchers, sales leads for the enterprise tier, public sector specialists, and policy staff to handle the EU AI Act rollout. Both firms are also building out recruiting, finance, legal and HR back-office teams in London, because you cannot run a 500 to 800 person office without them. For mid-career professionals outside pure research, the operational and GTM roles often have the widest funnel and the fastest hiring cycle.

The Salary Picture Is Aggressive

London compensation for AI talent has decoupled from the rest of UK tech. According to Sifted, Anthropic is currently advertising machine learning research engineer roles in London with base salaries up to £630K, which translates to roughly $780K at current exchange rates. Reinforcement learning research engineers are listed at £225K to £340K base, while security engineers focused on AI threat monitoring sit at £255K to £325K. OpenAI's London bands are not all public, but US research engineer ranges of $295K to $530K base are a reasonable proxy, and total comp lifts substantially with equity.

For context, UK senior software engineer median pay outside of frontier AI sits around £75K to £110K. The Anthropic and OpenAI bands are running three to six times that. The knock-on effect is already visible. Recruiters report frontier AI labs are pulling senior engineers out of fintech, hedge funds and even US-headquartered scaleups with London branches. If you have been undervalued at your current firm, this is a market reset moment.

How International Talent Can Actually Get In

Visa friction has historically been the choke point for US and Asian candidates eyeing London roles. Two routes have collapsed that timeline.

UK Global Talent Visa

The Global Talent visa is unsponsored, requires no job offer, and covers AI specialists, software engineers, fintech and cybersecurity professionals. Endorsement comes through Tech Nation's successor body or the Royal Society for research candidates. Approved applicants can switch employers, freelance, or start a company, which is uniquely flexible compared to the standard Skilled Worker route. The Global Talent endorsement timeline runs four to eight weeks for most applicants, and a new design route is launching in July 2026 to broaden eligibility further.

High Potential Individual Visa

The High Potential Individual visa is the fastest unsponsored path for recent graduates of top global universities. The 2026 list includes most US Ivy League schools, top UK and European institutions, and select Asian universities. Bachelor's and master's holders get two years of UK work rights, and PhD holders get three years. From November 2025, the route is capped at 8,000 applicants annually, and from January 2026 the English language requirement rose to B2 level. The HPI visa does not require sponsorship or even an offer in hand, which is why frontier AI labs love it. They can hire a Stanford ML PhD without involving immigration counsel, and the candidate can land in London inside a month.

For US-based candidates considering relocation, the math is straightforward. London compensation is now competitive on a take-home basis once you factor in lower healthcare costs, denser AI peer networks, and equity that is denominated in dollars. The UK income tax burden is real, but base salaries at Anthropic London are high enough to absorb it.

Why London, And Why Now

The UK government has been aggressive. The AI Opportunities Action Plan, launched in January 2025 and refreshed in early 2026, set out a 50-point program to make Britain the third global AI hub after the US and China. Key components include an internal headhunting capability inside government to bring elite individuals to the UK, expansion of compute infrastructure, and dedicated AI growth zones. The plan explicitly calls out London's Knowledge Quarter as core infrastructure.

DeepMind has anchored the cluster since Google acquired it in 2014. Around the corner sit Meta's London AI team, Wayve's autonomous driving researchers, Synthesia's generative video team, and a long bench of academic spinouts from Oxford, Cambridge and Imperial College London. The talent density is now self-reinforcing. A new lab opening in London today inherits a pre-built recruiting funnel of 10,000 plus AI-trained engineers within commuting distance. That is why Bloomberg framed the leasing wave as a structural shift, not a one-off real estate trade.

For job seekers tracking these openings in real time, Metaintro surfaces postings as soon as labs publish them, so you can move before the recruiter inbox fills up.

What This Means For UK And US Job Seekers

If you are UK-based with a CS or ML background, your near-term leverage is at a peak. Both labs are running structured interview loops with three to five rounds, technical screens, system design or research design, and behavioral panels. Standard prep applies. Brush up on transformers, attention mechanisms, RLHF and constitutional AI methods for Anthropic specifically. For GTM and operations roles, the emphasis is on enterprise complexity, regulated industry experience, and EMEA market knowledge.

If you are US-based and considering a move, the tradeoff math has shifted. The visa friction is low, the compensation is competitive, the cluster is real, and the cost of living gap with the Bay Area has narrowed. For mid-career engineers tired of the San Francisco grind, London offers a denser AI peer network than New York or Seattle, with shorter commutes than the Bay.

The third group is the one most underrated, the career switchers from finance, consulting and traditional tech. Both Anthropic and OpenAI are building out solutions, sales, policy, communications and recruiting teams that do not require a research background. These roles often pay £150K to £300K total comp in London, and the interview bar is rigorous but not gated by a PhD.

How To Prepare A London AI Application That Lands

Recruiters at frontier labs sift hundreds of inbound applications per role, so the candidates who break through tend to share three traits. First, a public artifact that proves you can do the work. For research roles that means a paper, a model card, an open-source contribution to a transformer library, or a write-up of a non-trivial fine-tuning project. For GTM roles it means a documented enterprise deal, a deployment case study, or a published point of view on AI buyer behavior. Second, a genuine point of view on safety, alignment or applied AI use cases, written down somewhere a recruiter can find it. Anthropic in particular reads applicant blogs and Substacks. Third, a tight one-page resume that leads with quantified outcomes rather than tool lists. The labs already know you can use Python and PyTorch.

For interview prep, both companies share materials publicly. Anthropic's career site walks through their interview structure, and OpenAI's recruiting team has posted research interview guides on the arXiv ecosystem. Practice live coding on a shared editor rather than in your IDE, because the screen-share format trips up candidates who normally rely on autocomplete. For research design rounds, expect open-ended prompts about scaling laws, evaluation methodology, or how you would investigate a specific failure mode in a frontier model. There is no single right answer. Recruiters are watching how you reason out loud, scope the problem, and identify what data you would need to be more certain.

Three Moves To Make This Week

If London is on your radar, three concrete actions move the needle in the next seven days. First, audit your professional profile and resume against the actual job descriptions on the Anthropic and OpenAI careers pages, not against a generic AI engineer template. The keyword match between resume and posting still drives most applicant tracking system rankings. Second, if you need a visa, start the Global Talent endorsement application now rather than waiting for an offer. The four to eight week endorsement window is the slowest part of the timeline, and you can submit it before any conversation with a recruiter. Third, warm up two or three London-based engineers or researchers in your network for a coffee chat. Internal referrals at frontier labs convert at multiples of cold applications, and the cluster is small enough that two degrees of separation is the norm.

For job seekers anywhere in the world, the takeaway is that the AI hiring map is no longer Bay Area first and everywhere else second. London has just locked in a 1300 plus seat surge across the two most discussed labs in the industry, and the surrounding cluster of DeepMind, Meta, Wayve, Synthesia and Isomorphic Labs guarantees that the next round of openings will follow. Watching this market closely, applying selectively, and preparing properly is the playbook for the next twelve months.

People Also Asked

Q: How many jobs are Anthropic and OpenAI opening in London?

A: Anthropic's new 158K square foot office holds 800 people, against a current UK headcount around 200, implying roughly 600 net new roles. OpenAI's 88,500 square foot lease seats 500-plus, against around 200 currently in London, implying about 300 net new roles. Combined, that is roughly 900 new openings inside two leases, with onboarding scaling through 2026 and into early 2027.

Q: What visa should I use to move to London for an AI job?

A: The UK Global Talent visa is the most flexible unsponsored route for experienced AI professionals, with endorsement timelines of four to eight weeks. The High Potential Individual visa is faster for recent graduates of qualifying global universities, with no sponsorship or job offer required. Both let you start interviewing before you have an employer locked in.

Q: Are London AI salaries competitive with US offers?

A: Yes for top research roles. Anthropic London is advertising machine learning research engineer base salaries up to £630K, which is roughly $780K. Reinforcement learning roles run £225K to £340K base, and security engineers £255K to £325K. Total comp including equity often matches or exceeds Bay Area packages on a cost-of-living adjusted basis, though headline US numbers can still be higher for the very top tier.

Ready to explore?

Metaintro tracks every major hiring wave in real time, from frontier AI labs to enterprise tech and beyond, so you can spot the right window before it closes. Sign up to get matched to roles that fit your background, location preference and visa status, and start your London move with a shortlist of openings that actually want to talk to you.