The UK government is moving to position London as a strategic sanctuary for AI giant Anthropic, attempting to capitalize on a deepening rift between the San Francisco-based startup and the US defense establishment. As the company faces a “supply-chain risk” designation from the Pentagon and public condemnation from Donald Trump, British officials are sketching out a high-stakes proposal to lure the $380bn (£287.4bn) firm into a deeper commitment to the UK, ranging from a major office expansion to a potential dual listing on the London Stock Exchange.

The strategy, led by the Department for Science, Innovation and Technology (DSIT) with support from No 10, comes at a moment of extreme vulnerability for Anthropic in its home market. The US government blacklisted the company after Anthropic refused to allow its Claude chatbot to be used for US surveillance or autonomous weapons—a move that ended a lucrative Pentagon contract. While a US judge has temporarily blocked the blacklisting and a second lawsuit regarding the supply-chain risk designation is pending, the relationship with Washington has turned toxic.

The friction in the US has been amplified by political rhetoric. President Donald Trump recently criticized the firm’s employees as “leftwing nut jobs,” posting on Truth Social in February that the US would not allow a “radical left, woke company” to dictate military strategy. This hostile environment has created a window for the UK to present itself as a “steadfast” and “pro-innovation” alternative.

The London Pivot

London Mayor Sadiq Khan has already stepped into the fray, writing to Anthropic CEO Dario Amodei to invite the firm to establish a more permanent base in the city. The effort is not merely about prestige; it is a targeted attempt to reduce the UK’s reliance on foreign AI while securing a “large-player” for a London Stock Exchange that has struggled to attract top-tier AI listings.

The partnership is already yielding tangible results. Anthropic has partnered with the UK government to integrate AI assistance into GOV.UK services, signaling a level of institutional trust that contrasts sharply with its current standing in Washington. The firm already employs around 200 people in Britain and took a significant step in local political alignment last year by appointing former prime minister Rishi Sunak as a senior adviser.

The culmination of these efforts will likely occur in late May, when Amodei visits the UK to meet with policymakers and European customers. During this visit, DSIT is expected to present its formal proposals to the CEO.

Closing the AI Capability Gap

For the UK government, wooing Anthropic is part of a broader, more urgent mission to build domestic AI capabilities. Having realized it lacks a homegrown competitor to the leading US labs, the UK recently unveiled plans for a £40m state-backed AI research lab.

The UK is effectively competing for the “brain trust” of the AI era. OpenAI has already committed to making London its largest research hub outside the US, and Google continues to anchor its presence through the £1bn King’s Cross site and its acquisition of DeepMind. By pursuing a dual listing for Anthropic, Business Secretary Peter Kyle aims to lock the company into the UK’s financial ecosystem, providing the LSE with the AI prowess it currently lacks.

Will Anthropic actually dual-list in London?

While a dual listing is described by some officials as “the dream,” it remains speculative. Anthropic is already in talks with major Wall Street banks for an IPO. A dual listing would require the company to navigate two different regulatory environments and investor bases, but it could provide a strategic hedge against continued political volatility in the US.

How did the conflict with the US Pentagon start?

The clash centered on “red lines” regarding warfighting. Anthropic refused to permit the use of its technology for autonomous weapons or surveillance, leading the US defense department to designate the company as a supply-chain risk and terminate its Pentagon contracts.

What is the immediate impact on UK citizens?

The most direct impact is the integration of Anthropic’s AI assistance into GOV.UK services, which is intended to streamline how citizens interact with government digital portals.

Why does a $380bn valuation matter for the UK?

Attracting a company of this scale would be a massive validation of the UK’s post-Brexit financial strategy. Beyond the tax revenue and jobs, it would signal to other global tech firms that London is a viable, stable alternative to Silicon Valley for high-valuation AI enterprises.

Can the UK offer enough regulatory stability and capital incentive to convince a US giant to split its loyalty between two continents?

The “swipe economy” is hitting a wall. For a decade, the dominant business model for digital romance was built on the premise of infinite optionality—a digital catalog of humans where more choices equaled better odds. But by the final quarter of 2025, the cracks became financial realities: industry giants Tinder and Bumble both reported a decline in paying users, even as the cost to retain those users climbed. In the tech epicenter of San Francisco, a new, more desperate market is emerging, shifting away from the infinite scroll and toward “agentic” matchmaking, high-stakes financial bounties, and a sudden, urgent return to organic human interaction.

The failure of the traditional app model is best illustrated by the “infinite indifference” now plaguing Gen Z. When the pool of potential partners is reduced to a game of rapid-fire visual judgment, the result is often a series of “situationships” rather than stable partnerships. In San Francisco, where the gender imbalance is notoriously skewed toward young men, the inefficiency of these apps has led some users to treat dating like a gig economy transaction.

Patricia Tani, cofounder of the AI-human gig platform RentAHuman, bypassed the apps entirely by posting a $200 “bounty” on her company website for a Valentine’s Day date. Her requirements were specific: a “sigma nerd” with “rizz.” The result—200 applications—suggests a significant appetite for high-intent, transactional entry points into dating, providing a level of filtering that traditional algorithms have failed to deliver.

From Catalogs to Curators: The Rise of the AI Matchmaker

A new guard of startups is betting that the cure for technology-induced burnout is not less technology, but more sophisticated, “agentic” versions of it. Companies like Known, which raised nearly $10 million last year, are attempting to scale the intimacy of a professional matchmaker. Instead of a profile, Known uses a voice-based AI to conduct a minimum 10-minute interview, probing for values and past relationship failures rather than superficial interests.

The commercial strategy here is a deliberate contraction. Whereas Tinder and Bumble thrive on keeping users in the app, Known delivers only one match at a time. By shrinking the perceived market—estimating the active San Francisco dating pool at roughly 60,000 people—the app attempts to replace the “infinite scroll” with a curated introduction. For a $15 fee, users can unlock the identity of their single, AI-selected match.

Other players are pursuing similar “swipeless” futures. London-based Fate and college-centric Ditto leverage voice AI and weekly limits to engineer scarcity. However, the efficacy of these models remains contested. Some users report that AI-driven matches lack the “je ne sais quoi” of human chemistry, and others have noted a disturbing trend: matches using AI prompts to handle the initial conversation, creating a loop where two algorithms effectively date each other before the humans ever meet.

The High-Net-Worth Moat and the Bounty Model

While AI attempts to democratize matchmaking, the ultra-luxury segment remains a human-led fortress. Professional matchmaker Blaine Anderson, whose packages can reach $100,000, argues that chemistry cannot be codified. Her experience attempting to automate her expertise in Austin, Texas, ended in failure; the algorithm could find a profile that matched a client’s desires, but it could not ensure the other person felt the same.

To bridge the gap between $100k bespoke services and $15 AI matches, Anderson launched “Bring Me Bae.” The platform leverages the “Date Me Doc” trend—essentially a dating résumé—and adds a financial incentive. Singles offer rewards between $10,000 and $30,000 to anyone who introduces them to a partner, with funds held in escrow and paid out only after a year of successful relationship. This model treats the “find” as a high-value acquisition, signaling “high intent” and filtering out the casual users who dilute the experience on mass-market apps.

The Agency Paradox

There is a profound irony at the center of the San Francisco scene: the more “agentic” the technology becomes, the less agency the users possess. The rise of “rizz keyboards”—AI tools that draft messages for users—suggests a generation that is increasingly ill-equipped for the friction of real-world rejection. We are seeing a market for tools that simulate confidence and charisma given that the digital environment has atrophied the actual skills required to sustain a relationship.

This atrophy is driving a counter-trend toward “organic” channels. From singles events at Trader Joe’s to strict 50/50 gender-ratio parties at the San Francisco Mint, there is a visible yearning for liberation from screens. Young professionals who spent their formative years on Zoom are now treating in-person interaction as a luxury good.

the business of dating is shifting from the quantity of connections to the quality of the filter. Whether through a $10,000 bounty or a voice-AI interview, the goal is no longer to see everyone, but to be told exactly who to see.

Why are traditional dating apps seeing a decline in paying users?

The decline is largely attributed to “swipe fatigue” and the failure of the “infinite choice” model. As users encounter a surplus of low-quality matches and “situationships,” the perceived value of a paid subscription drops. The cost of user acquisition and retention is rising because the core product—the swipe—has develop into associated with indifference rather than discovery.

How does “agentic” matchmaking differ from traditional algorithms?

Traditional algorithms typically use collaborative filtering (people who liked X also liked Y) or basic demographic filters. Agentic matchmaking uses AI to act as a representative. It collects deep, qualitative data through interviews and makes a definitive decision on behalf of the user, removing the burden of choice and the inefficiency of the swipe.

What are the commercial implications of the “bounty” dating model?

The bounty model shifts dating from a subscription-based utility to a performance-based reward system. By putting significant capital in escrow, users signal extreme intent, which attracts a higher caliber of “headhunters” or introducers. This could potentially create a new niche market for high-end, incentivized matchmaking that bypasses the app ecosystem entirely.

Can AI actually solve the problem of romantic chemistry?

Current evidence suggests that while AI can optimize for “paper compatibility” (values, goals, demographics), it cannot yet predict the visceral, non-quantifiable elements of chemistry, such as banter or physical tension. This remains the primary competitive moat for high-end human matchmakers.

As we outsource the search for love to agents and algorithms, are we optimizing for the right partner, or simply optimizing for the avoidance of rejection?