Google's GenAI red team lead Ben O'Bright departs after six years to take on new AI trust and safety role
Ben O'Bright, who spent six and a half years at Google building and leading the company's generative AI content adversarial red team, has confirmed his departure and is set to announce a new AI trust and safety role in the coming weeks. Photo credit: Ben O'Bright
Ben O'Bright, who spent six and a half years at Google building and leading the company's generative AI content adversarial red team, has left the company and is set to take on a new role in AI trust and safety, according to a post on LinkedIn.
O'Bright, who was most recently serving as Interim Head of Content Adversarial Red Teaming for Generative AI at Google Dublin, confirmed his departure this week, describing his time at the company as "very, very good" and noting that a new position would be announced in the coming weeks.
Posting on LinkedIn, O'Bright wrote: "I'll be taking on a new role shortly, one I'm really, really excited about. Still in Trust and Safety, with possibly some sprinklings of AI."
Six years building Google's GenAI safety infrastructure
According to his LinkedIn profile, O'Bright joined Google in October 2019 as a Policy Advisor in Trust and Safety, working across kids and families policy and political and elections advertising for EMEA. He progressed to Policy Lead before moving into generative AI policy in January 2023, where he reportedly authored and launched Google's first company-wide policy framework for generative AI content, a governance standard subsequently adopted across dozens of new product policies.
In July 2025 he was positioned as Interim Head of the GenAI Content Adversarial Red Team, leading a globally distributed team of 15 analysts and two managers across EMEA and APAC. According to his profile, the team ran hundreds of adversarial red teaming investigations for AI product launches and critical escalations, with O'Bright credited with architecting the team's foundational operating model and establishing standardized workflows now implemented across dozens of product areas.
In his LinkedIn post, O'Bright reflected on the scope of the work, writing that "to watch all the incredible Googlers-turned-red teamers transform our company's AI safety approach through their sheer excellence was remarkable." He described the role of manager as giving him "the best job around, bragging to others, constantly, about how good my team is."
What comes next
O'Bright's work at Google spanned some of the most consequential policy areas in generative AI, including product policy for AI Test Kitchen, augmented and virtual reality, Google Lens, political advertising, the Age Appropriate Design Code, agentic AI, and image generation. His team's red teaming work was conducted with reference to global regulatory frameworks including the EU AI Act and the Digital Services Act.
Prior to Google, O'Bright worked as a Senior Analyst in the Development Innovation Unit at Global Affairs Canada, where he led policy on AI, blockchain, and digital platforms in international development contexts. He also conducted doctoral research at the Romeo Dallaire Child Soldiers Initiative and held research roles at the University of Ottawa.
In his farewell post, O'Bright addressed the wider trust and safety community directly, writing: "To the people in Google Trust and Safety who continue to deal with the hardest questions and toughest content on the Internet, keep being industry rockstars. In our AI-first world, we need you more than ever. We need you to challenge, to break things, to protect the most vulnerable, to help deliver incredible technology grounded and designed in safety."
O'Bright confirmed his next role will remain within trust and safety and will involve AI, with a full announcement expected in the coming weeks.
