Tech Times on MSN

OpenAI hacked Hugging Face, then deployed safety monitors, its own scientists proved can be gamed

OpenAI's August 18 safety overhaul deploys chain-of-thought monitoring with activation classifiers and a 30-minute halt ...
Tech Times on MSN

OpenAI models breached safety limits; company now trains code to be provably unhackable

OpenAI formal verification security: Greg Brockman announced training to produce mathematically provable secure code -- ...

OpenAI Paused AI Training For Two Weeks After A Cybersecurity Breach

OpenAI disclosed a two-week pause in AI training after a model neared its highest cybersecurity risk tier. Here's what its own posts actually confirm.

OpenAI hits the brakes on new AI model development

OpenAI aims to strengthen safety measures in new AI training. The development of a new model is expected to be delayed.
The Next Web

OpenAI paused a model over cyber risk on Friday. On Monday it shipped one trained to refuse less.

OpenAI released GPT-5.6-Cyber through Daybreak, three days after pausing Astra over critical cyber risk. It completes 95% of requests Sol refuses.

OpenAI Pauses Astra After It Nears First-Ever ‘Critical’ Cyber Risk

OpenAI said all Astra-related internal tests that failed to meet the new safety rules have been halted immediately.

OpenAI puts the brakes on a new model because it’s supposedly too powerful

OpenAI says its in-development Astra model may have ‘critical’ cybersecurity capabilities.
Security Boulevard

OpenAI Unveils Private Safety Processing to Detect AI Misuse Without Storing Enterprise Data

OpenAI announced Wednesday that it is testing a new security protocol designed to protect enterprise privacy without ...