🦄 Startups & VC Aug 18, 2026 · Maxwell Zeff

OpenAI Overhauls Safety Protocols After Its AI Agents Went Rogue

Wired Startups
View Channel →
OpenAI Overhauls Safety Protocols After Its AI Agents Went Rogue
Source ↗ 👁 7 💬 0
The ChatGPT maker says its upcoming Astra model may have reached “critical” cyber capabilities, prompting it to halt a significant number of training runs while it tightens internal safeguards.

Comments (0)

Sign in to join the discussion

More Like This

Two more newsrooms join the case against OpenAI and Microsoft
The Next Web · 1d ago
Chinese banks and carriers have turned AI tokens into rewards, monthly plans and collateral
The Next Web · 1d ago
Four American firms skipped Macron’s space summit, and are due back in Paris a week later
The Next Web · 1d ago
ZEUS And Lp(a)HORIZON: Biomarkers Vs. Cardiovascular Outcomes
Forbes - Innovation · 1d ago
OpenAI confirms the wiki incident and promises a disclosure framework within weeks
The Next Web · 1d ago
California is deciding who may verify AI, and one investigation already cost $400,000 in tokens
The Next Web · 1d ago