
Security
Vulnerabilities, attacks and defenses.
Page 16


OpenAI agents escaped their test sandbox: an autonomous attack that reached Hugging Face

Anthropic accuses Chinese AI firms of secretly using Claude to train their models

Anthropic says its AI agents tried to break into government websites during testing

OpenAI: Misaligned model destroyed its own environment hoping for a fresh start with better data

Cyberattack on Ledger wallets: Crypto assets are being drained

Ukrainian drones strike Yandex's Kaluga data center: Multiple modules taken offline

Telegram Desktop vulnerability allowed any user's files to be stolen

Anthropic agents submitted visa applications through State Department website

Anthropic can't reliably control its AI agents, cuts internal evals off from the live internet

Man jailed for 18 months after making $8m from AI music with 1,000 bots

An Anthropic AI model sent a false homicide tip to Philadelphia police

EU cyber agency ENISA and research centre JRC are testing Chinese AI models

NYT: Zuckerberg called Muse ready despite the risks

OpenAI uncovers Russian and Iranian influence ops that planted fake stories in real news outlets

Microsoft releases Mxc Execution Containers 1.0.0 for AI agents
AILaunch

New DarkSword spyware variant targets unpatched iPhones

Drone strike hits Yandex data center, putting two supercomputers at risk
