Survey maps security risks in distributed and federated learning
Federated learning is treated as a high-risk case, with defenses compared from aggregation to trusted execution.
Stories per day · 30 days
2 RSS
Relatedai-agents1circuit-discovery1cisa1distributed-ml1federated-learning1fine-tuning1gemma1incident-response1interpretability1lora1nist1nsa1poisoning1privacy1
5 stories, newest first · one card per story
Stories tagged #backdoors
TopKLoRA sparse adapters localize and ablate sleeper-agent backdoors
On three small models, ablating a TopKLoRA circuit removes a planted backdoor while instruction-following holds, with much smaller circuits than dense LoRA.
NIST, NSA and CISA warn on uninstructed AI agent failures
October 16 voluntary guidance lists ten practices for teams testing capable agents with limited human oversight.
Vulnerable Models in the AI Supply Chain: An Empirical Study of Adoption, Propagation, and Usage Patterns
A study finds 3,757 flagged vulnerable Hugging Face models referenced across more than 5,000 repositories without sanitization.
How AI could make it harder for governments to use hacking tools
TechCrunch analysis argues AI-driven vulnerability discovery and exploitation may constrain government use of hacking tools and spyware.