Google DeepMind Is Preparing For AI’s Biggest Risk. Biosecurity Is Becoming A New Battleground.

Google DeepMind has launched a new bioresilience program aimed at helping governments, researchers and biosecurity organizations detect and respond to biological threats as concerns over the intersection of advanced artificial intelligence and biotechnology continue to gain attention.

The initiative, announced Thursday, brings together Google DeepMind and drug discovery company Isomorphic Labs to support efforts in pathogen surveillance, vaccine and therapeutic design and outbreak response. The company said it has established more than 15 partnerships with governments, researchers and biosecurity groups over the past year.

The announcement comes amid broader discussions around the safety and governance of advanced AI systems. Earlier this week, DeepMind Chief Executive Demis Hassabis called for governments to establish an international standards body for frontier AI models, arguing for a more structured approach to oversight. The proposal follows growing debate in Washington and other capitals over the potential national security implications of increasingly capable AI systems. Axios reported that Hassabis has advocated for a framework funded by industry but accountable to the U.S. government.

Google said it is also expanding access to some of its AI models and agents to trusted governments, researchers and biosecurity partners. The company classifies these deployments as "low-risk" restricted releases under its existing safety framework, according to Axios.

"If we were to find that we were reaching a critical capability level and we didn't have the appropriate mitigations, then we would not be launching," Helen King, vice president of responsibility at Google DeepMind, told Axios, adding that the company conducts ongoing evaluations throughout model development.

The focus on biosecurity comes as governments increasingly examine the risks associated with advanced AI systems. Earlier this year, the White House released an executive order directing federal agencies to establish testing and evaluation standards for frontier AI technologies, while lawmakers have continued to debate safeguards related to national security and critical infrastructure, according to Reuters.

Google has spent years building tools at the intersection of biology and computing. Its AlphaFold system, which predicts protein structures, has been widely adopted by scientists studying diseases and drug development. Hassabis and DeepMind researcher John Jumper received the 2024 Nobel Prize in Chemistry for their work on AlphaFold, according to the Nobel Prize organization.

Meanwhile, Isomorphic Labs has continued to expand its drug discovery efforts. The company announced earlier this year that it had entered additional partnerships with pharmaceutical companies to accelerate the development of treatments using AI-assisted research tools, according to Reuters.

Industry attention on biological applications has increased alongside improvements in AI capabilities. In a recent report, the nonprofit research organization RAND highlighted both the opportunities and risks associated with using advanced AI systems in biotechnology, noting that safeguards and controlled access remain key considerations for governments and private companies.