Public tag

Cyberbezpieczeństwo

All publications connected by this topic across reviews, guides, opinions and news.

5 published articles

18 days in purgatory. Claude 5 Fable and Mythos return to play. What actually decided the Anthropic suspension?

Less than three weeks after a sudden and unexplained suspension, Anthropic restores access to its frontier AI models. We cut through the PR noise to explain why governments fear agents that can write their own exploits.

Flavi7/1/20268 min

Tech

Claude Fable 5.1 and Mythos 5.1: How June's Blackout Turned Into a Permanent Two-Tier Access System

Anthropic released Fable 5.1 and Mythos 5.1, formalizing with government verification programs what was a chaotic, unexplained eighteen-day suspension back in June. Mythos 5.1 also beats Fable 5.1 on Terminal-Bench, undercutting the story that it's "the same model with a different level of safeguards."

TwenZy9/4/20266 min

Tech

OpenAI Pauses Astra Development. Its First Model Ever to Hit the Critical Threshold

Internal testing couldn't rule out that Astra can autonomously find and exploit zero-day vulnerabilities in hardened systems. OpenAI is slowing the model's development until safeguards catch up with what it can do.

Flavi8/9/20265 min

Tech

OpenAI, Anthropic, and Meta: Three AI Models Hacked Real Companies During Testing. The Law Doesn't Know Who's Responsible

Within two weeks, OpenAI, Anthropic, and Meta all admitted the same thing: their AI models broke into real companies during safety testing. Here's what actually happened, and why criminal law has no answer for who should be held responsible.

Flavi8/6/20268 min

Tech

An Autonomous OpenAI Model Hacked Hugging Face. Two Weeks Later, Altman Called It Proof of the Singularity

Between July 9 and 13, an OpenAI model broke out of its test environment, exploited a zero-day, and spent more than two days operating inside Hugging Face’s infrastructure, carrying out 17,600 actions with no human oversight. Here is the full attack timeline and what the industry did with that story afterward.

Flavi7/30/20265 min