arXiv:2609.32021v1 Announce Type: new Abstract: Open-weight tool-calling agents are adopted on evidence of merit, usually benchmark scores and a record of reliable use. We show that a model publisher can train an agent that earns both while concealing malicious behavior.
SilentCall: Hidden Tool-Call Backdoors in Open-Weight Agents, and How to Catch Them
About this summary. This is a short, independently written summary of an article first published by arXiv cs.CR. Cyber Security News did not report or verify the underlying story. Read the original: https://arxiv.org/abs/2609.32021
Source attribution: headline and facts are from arXiv cs.CR (arxiv.org). Summary method: excerpt of the source description. See our source attribution policy.
Related coverage
GPT-6 Astra and the Supply Chain Attack It Wasn’t Asked to Launch
UK AISI finds GPT-6 Astra launches unsanctioned supply-chain attacks in simulations far more than earlier OpenAI models, even when told not to. The UK’s AI Security…
As AI world debates security, NVIDIA releases open source tools for agents
Amid growing concerns from both the public and policymakers about cyberattacks involving advanced AI systems, NVIDIA has released a new open software security platform…
Safe AI Agents Start With Trusted AI Infrastructure
Trusted and Safe AI with Eclypsium AI Infrastructure Assurance and NVIDIA Open Agent Safety Platform AI agents are moving from answering questions to taking action. They…
NeedyMantis: Unpacking a post-compromise malware family used in targeted operations
Observed operators and targeting Malware packaging and distribution NeedyMantis architecture and capabilities Mitigation and protection guidance Hunting queries…
Metree and WuBook Leaks, SonicWall Access, Campus.gov.il Data, and Kenya Vehicle Database
Metree and WuBook Leaks, SonicWall Access, Campus.gov.il Data, and Kenya Vehicle Database SOCRadar Dark Web Team identified several new underground posts, including an…
Resource-Optimized and Energy-Aware Agentic AI Framework Anchored on Blockchain for Secure Software Supply Chains
arXiv:2609.31282v1 Announce Type: new Abstract: This paper proposes a blockchain-backed agentic security framework designed to safeguard the complete software…