Google is moving a team focused on the risks and societal impact of artificial intelligence out of Google DeepMind, the research lab developing its frontier AI models, according to an internal email viewed by the Journal.

The change in reporting structure-which follows a reorganization that saw DeepMind co-founder and Chief Executive Demis Hassabis step aside into a new chairman role-has prompted some members of the team to raise concerns that its work will be negatively affected, two current DeepMind employees said.

Starting next week, the roughly 90-person "AI responsibility" unit will join Google's global affairs organization, which oversees lobbying and public policy for the company as a whole.

The reassignment is part of broader organizational changes that have the effect of making DeepMind-which Google acquired in 2014 and has functioned as a semiautonomous operation-more tightly integrated with the rest of the company. In his new role, Hassabis will focus on external engagements and big-picture topics around AI. Koray Kavukcuoglu, Hassabis's successor as head of DeepMind, will have the title of senior vice president rather than CEO.

"GDM is moving into a more typical PA setup with central teams in central Google," Helen King, a vice president at Google DeepMind who runs the responsibility team, wrote in an email to the team about the changes. PA refers to product area, a term Google uses for divisions such as Google Search. Other teams including HR and policy are also moving out of DeepMind and into central Google.

Time magazine earlier reported that Google was considering reorganizing some of DeepMind's safety teams.

Employees on the responsibility team have expressed concerns to managers in message board discussions and meetings that the move will hamper their ability to conduct independent research, just as scrutiny of the impacts of AI mounts amid a broader backlash to the technology.

The responsibility unit includes teams working on testing Google's AI models for chemical, biological, radiological, and nuclear risks and running studies about how people interact with AI and the psychological impacts of chatbots, according to the people.

Employees expressed worries that moving out of DeepMind would lead to less access and insight into what the teams developing Google's Gemini AI models are working on, hampering their ability to identify relevant research questions.

In a statement, a Google spokesperson said, "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."

The global affairs organization includes the central trust and safety team for Google. Other AI safety teams, including groups focused on model behavior, privacy and security, will remain within Google DeepMind.

King has sought to reassure her employees that the changes won't affect the team's day-to-day work and mission. The team's focus will remain on emerging and frontier topics in responsible AI-and sharing work externally through publications and events "continues to be supported and important," King wrote in the email. The team will retain access to DeepMind and keep its computing and head count resources, King wrote.

"Demis still cares a lot about Responsible AI and is still planning to be involved," King wrote.

The reorganization is scheduled to take effect at the beginning of September, the people said. Ahead of the change, some employees sought to move to research teams that will remain in DeepMind, but were denied, the people said.

In meetings about the changes, managers have acknowledged the risk that some DeepMind employees may seek positions at other AI labs as a result, the people said.

In recent months DeepMind has seen the departures of a string of top researchers as the company's flagship AI models have fallen behind those from Anthropic and OpenAI on capabilities such as coding. Earlier this month, Jeff Dean, a fixture of Google's AI efforts since its earliest days, left with three other senior DeepMind researchers to found a startup. Two of the company's most-prominent researchers, Noam Shazeer and John Jumper, left earlier in the summer for OpenAI and Anthropic, respectively.