AISI has announced the launch of a new research partnership with Google DeepMind, marking the next step in a collaboration that has spanned two years. The partnership builds on existing work on frontier model testing and is part of DSIT’s broader engagement with Google DeepMind. The research Memorandum of Understanding, announced today, will enable deeper two‑way collaboration, including shared access to data and ideas, joint publications, and dedicated work between the teams to advance the science of AI safety and security. AISI looks forward to building on the productive partnership and sharing the results of this expanded research with the wider community. The agreement is expected to strengthen collaboration between a leading AI safety organization and a top research lab, potentially accelerating progress in AI safety research and influencing industry standards.
Deepening our partnership with Google DeepMind
AISI has announced a new research Memorandum of Understanding with Google DeepMind, extending a two‑year collaboration that has focused on frontier model testing. The agreement will deepen two‑way cooperation, allowing shared data, joint publications, and dedicated work to advance AI safety and security. The partnership is part of DSIT’s broader engagement with DeepMind.
Why it matters
The agreement strengthens collaboration between a leading AI safety organization and a top research lab, potentially accelerating progress in AI safety research and influencing industry standards.
Topic dossiers Google DeepMind & Gemini