A new academic attack called BadTV embeds a persistent backdoor inside the reusable "skill files" that let AI models pick up or drop capabilities without costly retraining, triggering malicious behavior whether the skill is added, removed, or recombined. The work, accepted at the ACM Conference on Computer and Communications Security (CCS), targets a fast-growing shortcut in machine learning known as task arithmetic and warns that the flexibility it offers comes with an unaddressed security cost.
Continue reading
The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in✓ Signed in — this article isn’t included in your current plan.
Unlocking the full article...
Contact us
Questions, corrections, or ideas? Send us a message here. We aim to reply by email within 24 hours.
Thank you. Your message has been sent. We aim to reply by email within 24 hours.Please enter your email address and a message.We could not send your message. Please try again in a little while.