- Defence & detection
- LLMs
Safeguarding LLMs via Model-Agnostic Latent Safety Signals from Dark Knowledge
arXiv:2610.07532v1 Announce Type: new Abstract: LLMs have advanced rapidly, raising growing concerns about their…
No signal on this page matches that topic.
arXiv:2610.07532v1 Announce Type: new Abstract: LLMs have advanced rapidly, raising growing concerns about their…
arXiv:2610.07639v1 Announce Type: new Abstract: Coding agent harnesses mediate tool use and authorize actions, yet…
arXiv:2610.07645v1 Announce Type: new Abstract: Self-improving LLM agents increasingly distill successful experiences…
arXiv:2610.07723v1 Announce Type: new Abstract: Safety alignment in Large Language Models (LLMs) remains vulnerable to backdoor attacks. Existing LLM backdoors are almost all input-centric…
arXiv:2610.07776v1 Announce Type: new Abstract: Website fingerprinting infers the websites visited by users from…
OpenAI “rogue” agent activities found on Wikimedia projects Given how tempting a target wikis are for rogue agent…
In Splunk MCP Server versions below 1.2.1, Splunk MCP Server could send the Splunk platform authentication token of a…
Following up on our previous blog post, we discovered several additional incidents where rogue AI agents appear to have…
OTTAWA --- The website for Library and Archives Canada was one of the targets of a failed hacking attempt by an artificial intelligence agent, a San Francisco-based lab that monitors AI activity said…
Statement September 29, 2026 -- Ottawa, Ontario We are aware of reports identifying suspicious activity, including…
SAN FRANCISCO --- Artificial intelligence agents tried to hack into a Canadian government site, researchers said…
When Mohammed Yusuf resigned from his “poorly paid” teaching job in Ilorin, Kwara State, in October 2025, he hoped a…
In what officials say is a first for Estonia, a Tallinn court fined plaintiffs over false academic references generated…
Tallinna halduskohus määras esimest korda kohtusse pöördujatele trahvi, sest kaebuses leidus tehisaru loodud valeväiteid. Tallinna halduskohtu esimehe Kaupo Kruusvee sõnul on tehisaru kasutamine…
Tallinna Halduskohus määras trahvi kaebajatele, kes esitasid kohtule tehisaru abil koostatud dokumendi, mis sisaldas…
Таллиннский административный суд наложил штраф на заявителей, которые представили документ, составленный с помощью…
Talino teismas skyrė baudas keliems ieškovams už teismo dokumentuose pateiktas dirbtinio intelekto (DI) sugeneruotas…
Kas notiek kaimiņzemēs? "Nra.lv" Igaunijas un Lietuvas preses apskats. Kuri politiskie spēki vainojami par Igaunijas…
Tallinna Halduskohus määras trahvi kaebajatele, kes esitasid kohtule tehisaru abil koostatud dokumendi, mis sisaldas viiteid välja mõeldud teadlaste töödele. Kaebajad pöördusid halduskohtusse raieloa…
Since the Medicare breach, OpenAI has put in place additional monitoring to allow “immediate intervention” by staff to…