Results for “"model"”
71–80 of 1,074 articles
Best match · title matches first
The Register
4 months ago
Researcher poisons open-weight AI model for under $100
Models demand trust without offering verification
ransomware hospital — both words
"data breach" — exact phrase
malware -android — exclude a word
(apple OR microsoft) patch — either company, plus patch
ransom* — a wildcard match
Searches article titles, summaries, keywords and sources. Filters use publication dates.
71–80 of 1,074 articles
Best match · title matches first
The Register
Models demand trust without offering verification
The Register
If you want a picture of the future of LLM security, imagine Whac-a-Mole meets Groundhog Day
The Hacker News
A new, stealthy backdoor named Mistic has been deployed as part of suspected financially motivated attacks aimed at multiple organizations spanning insurance, education, IT, and professional services sectors since April 2026. According to Symantec and Carbon Black's Threat Hunter Team, the backdoor, also tracked as MLTBackdoor, is said to be linked to an initial access broker (IAB) named
The Register
PARTNER CONTENT Europe wants control over its own technology, but what does that look like?
The Hacker News
A flaw in the Google Cloud Vertex AI SDK for Python let an attacker with no access to a victim's project hijack the victim's machine learning model upload and run code inside Google's serving infrastructure. Palo Alto Networks Unit 42, which found and reported the bug through Google's bug bounty program, calls the technique "Pickle in the Middle" and said it saw no exploitation in the wild.
Dark Reading
Anthropic abruptly suspended all access to Fable 5 and Mythos 5 after receiving an export control directive that banned foreign nationals from using the AI models.
The Hacker News
University of Toronto researchers have built and tested a proof-of-concept AI-driven computer worm that uses a locally hosted open-weight large language model to reason its way through a network, generate tailored attack strategies for each target it encounters, and replicate itself, all without human intervention and without touching a commercial AI service. The preprint, posted to arXiv on
Dark Reading
The White House's executive order establishes voluntary framework for early government access to frontier models while investing in federal security.
The Register
'Attackers can now cheaply operationalize known vulnerabilities at scale,' boffins tell The Reg
Bleeping Computer
OpenAI says it's rolling out a new update that improves the existing GPT-5.5 Instant model, and this move comes ahead of the scheduled retirement of multiple legacy models, including o3. [...]