On August 5, 2026, Google announced that Demis Hassabis would transition from his role as CEO of DeepMind to take on a broader position within Alphabet, focusing on the company's overall AI strategy. This move follows the 2023 merger of DeepMind with Google's internal AI lab, which created Google DeepMind, with Hassabis at its helm. This restructuring was a direct response to Google feeling "caught flat-footed" by the emergence of OpenAI's ChatGPT, aiming to streamline and strengthen its AI efforts.
The shift in Hassabis's role signifies a change in dynamic, as he now has increased power but potentially less direct control over individual units. For instance, he has had to adapt his stance on previously held "red lines." The company has also recently disbanded its dedicated AlphaFold team, reassigning researchers to focus on the Gemini large language model and other scientific projects, including enzyme design and genomics. This reallocation aims to tighten the connection between DeepMind's research and revenue-generating products like Google Cloud, and tools for healthcare and energy customers.
DeepMind's financial records show cumulative revenues exceeding $7.8 billion over the five years ending in 2024, though all of this revenue was internal, coming from payments for DeepMind technology used by other Google platforms. The company has also deployed over $9.6 billion of Google capital for operating expenses as Hassabis pursued artificial general intelligence (AGI). The team behind the Gemini app was also moved to Google DeepMind in October 2024 to improve feedback loops and accelerate model deployment.
This strategic evolution at DeepMind, including the reassignment of AlphaFold authors and the focus on Gemini-powered systems, reflects a shift from single scientific problems to developing frontier AI agents that can automate parts of the scientific process. While many AlphaFold authors have been reassigned or left the company, the development of AlphaFold will continue within broader scientific programs. This restructuring is part of Alphabet's larger effort to compete with rivals like Microsoft, OpenAI, and Meta in the race to supply general-purpose AI systems.
