October 9, 2025
It’s trivially easy to poison LLMs into spitting out gibberish, says Anthropic
Just 250 malicious training documents can poison a 13B parameter model - that's 0.00016% of a whole dataset Poisoning AI models might be way easier than previously thought if an Anthropic study is anything to go on. …
SonicWall breach hits every cloud backup customer after 5% claim goes up in smoke
Affects users regardless of when their backups were created SonicWall has admitted that all customers who used its cloud backup service to store firewall configuration files were affected by a cybersecurity incident first disclosed in mid-September, walking back earlier assurances that only a small fraction of users were…