Demis Hassabis was a personal, early investor in Anthropic, according to a report from the Financial Times, adding a new layer to the web of financial and ideological ties linking Google’s AI push to one of its sharpest rivals. The report said the Google DeepMind chief’s backing came while his company was building Gemini and Anthropic was racing ahead with Claude.
The disclosure matters because Google is already one of Anthropic’s biggest backers. The company has poured billions into the start-up, giving it an estimated $14 billion stake and major TPU deals, while Anthropic runs significant workloads on Google Cloud. That makes the relationship unusually close for two companies competing in frontier AI, even as they appear to share a view that the field needs guardrails.
Anthropic was founded in 2021 by Dario Amodei, Daniela Amodei and roughly a dozen other researchers who walked out of OpenAI over AI safety concerns. Amodei has described Hassabis as a role model, and the Financial Times report places the DeepMind co-founder inside the same safety-first orbit that has shaped Anthropic’s identity from the start.
Hassabis has been arguing against reckless open-sourcing of AI for years. As far back as 2016, he was making that case in an email to Elon Musk, who had co-founded OpenAI in part to prevent Google from controlling AGI. That history helps explain why the latest report landed with unusual force: it suggests the battle lines in AI are not just between companies, but between overlapping networks of investors, researchers and executives who may be rivals in public and aligned in private.
The tension is hard to miss. Google is building Gemini and backing Anthropic at the same time, while Anthropic is depending on Google infrastructure and capital even as it tries to prove itself as an independent challenger. For now, the new disclosure does not change the basic competitive picture, but it does show how quickly the frontier AI industry has become a place where the same people can be both competitor and patron.
