Google will merge its AI safety team with its DeepMind division as it ramps up its focus on AI development.
The Responsible AI team – tasked with Google’s AI safety plans – will be moved into DeepMind, the UK-founded AI company that became part of Google in 2014.
The consolidation will see the researchers from Responsible AI work directly with the teams building and scaling AI products such as Gemini. The development was first reported by Reuters.
As questions of regulation and AI safety become inescapable for tech firms, the move aims to create a more unified department that can efficiently incorporate safety research into development.
The London-headquartered Google subsidiary was a significant figure in the development of Gemini, a follow-up to Bard, an AI model created to compete with OpenAI’s ChatGPT.
It’s the latest restructuring of AI departments among tech giants. Last month, Microsoft appointed DeepMind co-founder Mustafa Suleyman as the CEO of its newly formed AI division. Suleyman is also the founder of Inflection AI.
Staff from DeepMind have also taken an active role in working with the government on AI safety.
In February, the AI Safety Institute – formed by the British government – announced Geoffrey Irving, who spent three years as a director at DeepMind, would join as a research lead for the institute.
Politico reported on Friday that Google DeepMind is the only major AI lab that has allowed the institute access to perform pre-deployment safety tests, a key pillar of Prime Minister Rishi Sunak’s AI safety deal announced last November.
The post Google to merge Responsible AI with DeepMind team appeared first on UKTN.
