You're tuned in to Tech Beat, and here's what's moving in technology today.
OpenAI's annual DevDay is underway in San Francisco, with CEO Sam Altman promising more than twenty launches. The company is teasing what Altman calls a new thing entirely, with speculation pointing toward a consumer AI agent designed to compete directly with Meta's recently unveiled Muse platform.
That excitement, though, comes with a shadow. OpenAI has reportedly pulled the planned public release of its GPT-six-point-one model after internal testing flagged safety concerns on two separate fronts. The company's own safety leadership confirmed the pause, and developers across the industry are now renewing calls for greater government oversight of frontier model releases.
Meanwhile, a quieter but troubling story out of the security research world. A team backed by Sequoia and Greenoaks has uncovered more than thirteen thousand sensitive corporate screenshots leaked through AI models, pulled from systems inside three hundred forty-three companies. The findings underscore a growing blind spot in how AI tools handle the data they encounter, and raise real questions about what it means to deploy these systems inside enterprise environments.
Keep surfing. Tech Beat out.
