Google Relocates AI Safety Unit Amid DeepMind Restructuring

Google has restructured its AI research lab DeepMind by relocating its AI safety unit to the company's global affairs organization. The move involves around 90 people focused on AI risks and impacts, which will now work closely with public policy and lobbying functions.

The shift aims to integrate AI safety efforts more closely with other areas of the business, reflecting Google's broader efforts to restructure DeepMind from a semi-autonomous lab to a more integrated division. This reorganization follows DeepMind CEO Demis Hassabis transitioning to a chairman role.

Some DeepMind employees are concerned that this change could limit their team's independent research and access to the latest AI developments. However, Google states that the move will enhance the ability of AI safety teams to inform model and product safety.