The AI responsibility unit includes 90 people and has teams focused on the chemical, biological, radiological and nuclear risks of Google’s AI models; the ways people interact with AI; and the psychological impacts of chatbots, according to the report.
The organization the unit is moving away from, Google DeepMind, is the company’s research lab developing frontier AI models, while the one it is joining, global affairs, oversees lobbying and public policy, per the report.
The move is part of Google’s effort to shift Google DeepMind from being the semiautonomous unit it has been since Google acquired it in 2014, to a division such as Google Search. It follows a reorganization that was announced Aug. 5 and saw DeepMind Co-founder and CEO Demis Hassabis move into a new chairman role, the report said.
Two unnamed current Google DeepMind employees told the WSJ that some members of the responsibility team are concerned that its work will be negatively affected by the change in reporting structure, that the teams’ abilities to conduct independent research will be hampered, and that the teams will have less insight into what’s happening with the latest models, per the report.
A Google spokesperson told the WSJ in a statement, per the report: “Many teams across Google work on AI safety and responsibility, and by bringing our AI responsibility teams closer together we’re strengthening their ability to inform safety for our models and products.”
It was reported in July that Hassabis called for a United States-led standards body to independently test the world’s most advanced AI models for national security threats before release. The proposal would cover frontier models developed in the U.S. and abroad, including open and closed systems.
PYMNTS reported in March 2024 that companies across the globe were intensifying their efforts to develop responsible AI systems, aiming to ensure fairness, transparency and accountability in AI actions, and that the process entailed appointing leaders in AI responsibility and training team members on ethical AI practices.
