Category
1,407 papers
This work determines what each reported result implies for that question: how much help with harmful tasks the service still gives an attacker who keeps adapting or finds another way in.
This work argues that red-teaming is better framed as a search problem: discover, organize, and iteratively refine a diverse archive of attack strategies, producing a structured map of how a target model fails rather than a list of one-off successes.
This paper claims that an in-depth investigation of the adversarial robustness of NeSy models is necessary and provides the first systematic evaluation of backdoor attacks against NeSy, and shows that while NeSy models are indeed more robust than their neural counterpart on average, their robustness vastly depend on th...
Deployed language model safeguards (safety fine-tuning, filtering, unlearning) vary by principal only outside the model weights: filters are reconfigured, tiers are multiplied, and artefacts are reissued; inside one set of weights every request meets the same model configuration. This motivates us to define capability-...
Reach audiences
Advertise in front of researchers, engineers, and readers.
It is shown that an external observer can identify the class of the workload running on an NVIDIA H200 from its power draw, and adversarial traces are recorded, offering initial insights beyond genuine activities and a dataset for developing and testing stronger evasion mechanisms.
It is argued that agent security must be evaluated under an untrusted-model assumption: a correct system is one in which a fully prompt-injected agent still cannot exceed the authority explicitly delegated to it, and an authorization broker is implemented that closes the gap.
This work formalizes four representative agent payment protocols: x402, MPP, ACP, and AP2 in Tamarin, and constructs source-grounded models that capture each protocol's roles, state, trust assumptions, and lifecycle transitions.
The theory that explains the blindness, the theory that explains the blindness, the two-population conservation law, under which every in-loop metric improves while true quality does not, and a synthetic study that validates the mechanism are given.
3R-Bench (Refusal, Repetition, and Revision), a benchmark of 150 real-world cybersecurity requests augmented with two adversarial conversational settings, is introduced and eight LLMs are evaluated, finding that prior assistant behavior strongly changes responses to an unchanged request.
OpenAgentFlow is presented, a control-plane/action-plane architecture that establishes the action-commit boundary as a shared enforcement interface and shows that a shared action-commit boundary provides a practical basis for system-wide governance across heterogeneous agent execution paths.
Gpt-oss-120b, a mixture-of-experts model with only $\sim$5.5B active parameters per token, carries the full state across all calls and returns the correct digest on a majority of completed runs, and localizes the residual failures by origin, separating state-carrying from arithmetic and from serving.
Experiments show that Knowledge Trap reduces surrogate Agreement by 6.2\% on average without degrading legitimate-user accuracy, outperforming existing defenses that impose measurable user impact, and suggest that defending knowledge-space traversal is a practical direction for mitigating LLM extraction attacks.
Research Manager Mike Walker reflects on a career spanning hacker culture, cybersecurity, and AI, sharing lessons about trust, teamwork, leadership, and why solving difficult problems requires curiosity, judgment, agility—and an understanding of the rules. The post Learn the rules: The hacker mindset appeared first on Microsoft Research.
GPT-Lab
· gpt-lab.eu
Oct 30, 2026
Critical flaws, compromised release channels and agentic AI threats landed within days of each other in mid-October 2026. Why “Black October” may say more about how well we now find weaknesses than about the attacks themselves. The post Black October? The Week AI Changed the Cybersecurity Conversation appeared first on GPT-Lab.
Gemini 4 Carbon is our strongest model with state-of-the-art cybersecurity and long horizon capabilities.
Gemini 3.8 Flash and 3.8 Flash Cyber deliver next-generation intelligence for agentic workflows and cybersecurity.
Google introduces Gemini 3.5 Flash Cyber, a lightweight cybersecurity model to find and patch vulnerabilities.
Students from the MIT Cybersecurity Clinic help local governments and other vulnerable organizations defend against digital threats.
We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.
