In a signal that the artificial intelligence gold rush is accelerating beyond even the most bullish forecasts, Anthropic, the San Francisco-based AI safety company, is reportedly on the brink of a public listing that could value it at nearly $1 trillion. The move has sent ripples through the global financial ecosystem, with London positioning itself as a secondary domicile for the company's shares — a bid to restore its reputation as a hub for high-growth tech listings.
Internal sources suggest Anthropic's board has been meeting with underwriters from Goldman Sachs and Morgan Stanley, finalising a dual listing that would see the firm trade on both the Nasdaq and the London Stock Exchange. If the $1tn valuation holds, it would make Anthropic one of the most valuable companies in history, overtaking Meta and trailing only Apple, Microsoft, and Saudi Aramco.
The figure is not arbitrary. Anthropic's large language models, particularly Claude, have become the de facto choice for enterprises seeking safety-focused AI. Unlike OpenAI, which has courted controversy with rapid deployments, Anthropic has positioned itself as the 'responsible' cousin — building 'constitutional AI' that aligns with human values. This ethical premium has attracted a raft of government contracts, including a recent deal with the UK's Cabinet Office for secure AI assistants.
But the real story is what a $1tn valuation says about the market's appetite for AI. We are not in a bubble, critics might argue. We are in a paradigm shift where the underlying technology is rewriting the rules of productivity. Anthropic's revenue, estimated at $4bn in 2024, is growing at 300% year on year. Compare that to the dot-com era's Pets.com, which had no revenue and a $300m valuation. This is different.
London's interest is particularly telling. Since the Brexit-induced exodus of tech IPOs, the LSE has struggled to lure high-growth companies. Arm chose New York. Deliveroo languished. But now, with regulators racing to create a 'crypto and AI-friendly' environment, the UK government has relaxed listing rules to allow dual-class share structures. Chancellor Rachel Reeves is said to have personally lobbied Anthropic's CEO Dario Amodei, promising tax breaks for 'digital sovereignty' research.
The subtext here is geopolitical. As the US-China tech war intensifies, the UK wants to be the 'Switzerland of AI' — neutral, trusted, and deeply integrated into Western supply chains. A London listing for Anthropic would be a powerful endorsement.
Yet the 'Black Mirror' consequences are impossible to ignore. A $1tn company built on algorithms we barely understand raises profound questions about power and control. Who audits Claude's decisions? What happens when a government contract for 'constitutional AI' interprets ethics differently than the electorate would? These are not academic concerns. They are the user experience of society in the 2020s.
Anthropic's own research shows that frontier models can already persuade humans in one-on-one conversations. A $1tn valuation means that the company's shareholders will have unprecedented influence over how that power is deployed. The company's 'long-termism' philosophy, which prioritises existential safety over short-term profits, is commendable. But financial markets are not known for patience.
For now, the narrative is one of triumph. London's fintech and AI clusters are buzzing. Goldman analysts are upgrading their price targets. But as someone who has watched the industry for two decades, I cannot shake the feeling that we are building a skyscraper on a fault line. The technology is miraculous, but the business model is untested. The regulation is nascent. And the valuation is built on faith that we can control what we have created.
Anthropic's float will be a watershed moment. It will either validate AI as the most transformative investment of our era, or it will be the pin that pops the balloon. Either way, the world will be watching from both sides of the Atlantic.
