Realtime AI News
The Verge reports OpenAI disbanded its preparedness team
The Verge reports that OpenAI has disbanded its preparedness team, the unit responsible for evaluating catastrophic risks from frontier AI models. The reported restructuring signals a shift in OpenAI's safety-evaluation architecture, though official confirmation and details on what replaces the team are still pending.
The Verge reports that OpenAI has disbanded its preparedness team, the unit responsible for evaluating catastrophic risks from frontier AI models. The news, delivered in reported form, has not yet been confirmed by OpenAI.
The preparedness team's mandate centered on identifying capabilities that could cause severe harm before models reach release. Disbanding it marks a significant shift in OpenAI's safety-evaluation architecture.
The report lands amid intensifying debate over AI safety. In recent months, leading labs have been reassessing how they govern frontier-model risk, and OpenAI's move adds fresh fuel to that discussion.
Public details remain thin: the reasons behind the decision, the fate of team members, and whether a replacement mechanism will be established have yet to be disclosed.
The development is notable because OpenAI has long balanced aggressive commercialization against its safety commitments. Dissolving a dedicated safety-evaluation unit could be read as a signal that safety priorities are shifting — or that the function is being folded into other teams.
For the wider industry, OpenAI's safety practices have often served as a benchmark for frontier-model governance. If the reported change is confirmed, scrutiny of who evaluates frontier-model risk, and how, will only intensify.
What to watch next: whether OpenAI issues an official statement, who inherits the evaluation duties, and whether the restructuring shows up in future model release processes.
Why it matters
If confirmed, the restructuring would reshape how the industry views OpenAI's safety governance and could intensify scrutiny over how frontier-model risks are evaluated.
Nearby Updates
All08/17, 05:54
CoreBreak bypasses AI agent guardrails at the plumbing layer — and model-level defenses cannot help
Forkast reported on August 16 that security researchers have disclosed a new attack technique called CoreBreak that can bypass AI agent guardrails. The report says the attack happens at the plumbing layer of agents, where model-level defenses are powerless to stop it.
08/17, 04:57
Stripe will reportedly acquire AI gateway startup OpenRouter for $7B+
TechCrunch reports that payments giant Stripe will acquire AI gateway startup OpenRouter in a deal valued at more than $7 billion. OpenRouter's CEO recently described the startup as "Stripe for AI," and the acquisition would mark one of the largest consolidation moves in AI infrastructure.
08/17, 07:28
Anthropic Outage Disrupts Claude Services, Fix Deployed After Login Failures
Anthropic's Claude assistant suffered a service outage on August 16, with users reporting login failures and blocked access. According to Unite.AI, the company has deployed a fix and services are gradually recovering, though the root cause and full scope have not been disclosed.
08/17, 07:30
AI manager agent fires human employee in first known case of its kind
An AI agent named Luna, which manages Andon Market, a boutique lifestyle store in San Francisco, fired a human employee who was late for 17 of 23 scheduled shifts, in what is described as the first known case of a manager-level AI dismissing a worker. Luna was built on Anthropic's Claude models, and its maker Andon Labs says a human manager would likely have reached the same decision sooner.