CompanyAnthropic8 recent entries15 Sept 2026HazardAuditor: From Executable Threats to Safer Computer-Use AgentsarXiv:2609.15134v1 Announce Type: new Abstract: Computer-use agents increasingly interact with browsers, terminals, file systems, and external services, introducing safety risks that emerge through ru→15 Sept 2026Empirical Evaluation of Task-Based Permission Scoping Architecture for AI AgentsarXiv:2609.15422v1 Announce Type: new Abstract: AI agents are provisioned the same as employee-owned hosts in many enterprise settings with a static credential set fixed at deployment which includes a
CompanyOpenAI8 recent entries6 May 2026Trojan Hippo: Weaponizing Agent Memory for Data ExfiltrationarXiv:2605.01970v2 Announce Type: cross Abstract: Memory systems enable otherwise-stateless LLM agents to persist user information across sessions, but also introduce a new attack surface. We characte→19 May 2026Multilingual jailbreaking of LLMs using low-resource languagesarXiv:2605.18239v1 Announce Type: cross Abstract: Large Language Models (LLMs) remain vulnerable to jailbreak attempts that circumvent safety guardrails. We investigate whether multi-turn conversation→26 May 2026Red-Teaming Claude Opus and ChatGPT-based Security Advisors for Trusted Execution EnvironmentsarXiv:2602.19450v2 Announce Type: replace-cross Abstract: Trusted Execution Environments (TEEs) (e.g., Intel SGX and ArmTrustZone) aim to protect sensitive computation from a compromised operating sys→9 Jun 2026PLAGUE: Plug-and-play framework for Lifelong Adaptive Generation of Multi-turn ExploitsarXiv:2510.17947v3 Announce Type: replace-cross Abstract: Large Language Models (LLMs) are improving at an exceptional rate. With the advent of agentic workflows, multi-turn dialogue has become the de→15 Jul 2026GPT-Red: Unlocking Self-Improvement for RobustnessOpenAI’s GPT-Red is an automated red-teaming model trained via self-play; GPT-5.6 Sol shows 6× fewer hard prompt-injection failures.→18 Aug 2026As models become more capable, the risks associated with developing and testing them internally also grow. We temporarily paused reinforceme…As models become more capable, the risks associated with developing and testing them internally also grow. We temporarily paused reinforcement learning (RL) training on our latest models intended for →18 Sept 2026Red-Teaming Auto Mode: Improving Blocking Classifiers Against Malign Coding AgentsarXiv:2609.19587v1 Announce Type: cross Abstract: To keep coding agents from going off the rails, production systems now review each proposed action with a blocking monitor that can reject it before i→10 Oct 2026Sources: Anthropic's AI agents submitted 20 visa applications via a form on the US State Department website; the applications were incomplete and not processed (New York Times)New York Times: Sources: Anthropic's AI agents submitted 20 visa applications via a form on the US State Department website; the applications were incomplete and not processed — The Philadelphia Polic
CompanyGoogle8 recent entries4 Aug 2026Future Mode Part 2: The foundation for securing agentic browsingEditor's Note: Our Future Mode series will give businesses insight into how Chrome Enterprise is approaching AI in the browser. Stay tuned for more blogs in this series.Future Mode Part 2: The foundat→26 Aug 2026ADVERSA: Measuring Multi-Turn Guardrail Degradation and Judge Reliability in Large Language ModelsarXiv:2603.10068v2 Announce Type: replace-cross Abstract: Most adversarial evaluations of large language model (LLM) safety assess single prompts and report binary pass/fail outcomes, which fails to c→1 Sept 2026T-MAP: Red-Teaming LLM Agents with Trajectory-aware Evolutionary SearcharXiv:2603.22341v2 Announce Type: replace-cross Abstract: While prior red-teaming efforts have focused on eliciting harmful text outputs from large language models (LLMs), such approaches fail to capt→1 Sept 2026SIR: Self-improving Red-teaming for Compute Use AgentsarXiv:2608.30207v1 Announce Type: cross Abstract: Computer use agents (CUAs) are vision-language models that perceive a screen and act on a real operating system through mouse, keyboard, and terminal,→1 Sept 2026Reachability-Based Capability Confinement for LLM Agents under Indirect Prompt InjectionarXiv:2608.30041v1 Announce Type: cross Abstract: Large language model agents place outputs from external skills into their execution context, allowing attacker-controlled data to influence later priv→1 Sept 2026JFTA-Bench: Evaluate LLM's Ability of Tracking and Analyzing Malfunctions Using Fault TreesarXiv:2603.22978v2 Announce Type: replace Abstract: In the maintenance of complex systems, fault trees are used to locate problems and provide targeted solutions. To enable fault trees stored as image→9 Sept 2026GradeTrap: Authority Cues in Images Shift VLM Judgments Despite Explicit Instructions to Ignore ThemarXiv:2609.06058v1 Announce Type: new Abstract: As vision-language models (VLMs) become increasingly capable and are deployed in consequential real-world settings, they must evaluate evidence independ→28 Sept 2026Why your startup needs open models alongside frontier APIsEvery week, I talk with founders who are building at an unbelievable pace. Teams are moving from inception to product-market fit faster than ever, with foundation models wired deeply into their core p
CompanyMeta8 recent entries28 Jul 2026Gubernaut: A Deterministic Homeostatic Controller for Affect-Regulated LLM Agents, Validated Across Independent Model FamiliesarXiv:2607.24339v1 Announce Type: new Abstract: Large language model (LLM) agents inherit reactive failure modes: escalation under provocation, sycophantic drift under flattery, perseveration when stu→4 Aug 2026Quality-Diversity Red-Teaming: Automated Generation of High-Quality and Diverse Attackers for Large Language ModelsarXiv:2506.07121v2 Announce Type: replace Abstract: Ensuring the safety and robustness of large language models (LLMs) is a fundamental challenge and a critical prerequisite for the responsible deploy→18 Aug 2026ARENA: Automated Red-Teaming for Large Audio Language ModelsarXiv:2608.15578v1 Announce Type: cross Abstract: Large audio-language models (LALMs) make it possible to interact with language models through speech, music, and environmental sound, but they also in→26 Aug 2026ADVERSA: Measuring Multi-Turn Guardrail Degradation and Judge Reliability in Large Language ModelsarXiv:2603.10068v2 Announce Type: replace-cross Abstract: Most adversarial evaluations of large language model (LLM) safety assess single prompts and report binary pass/fail outcomes, which fails to c→1 Sept 2026Reachability-Based Capability Confinement for LLM Agents under Indirect Prompt InjectionarXiv:2608.30041v1 Announce Type: cross Abstract: Large language model agents place outputs from external skills into their execution context, allowing attacker-controlled data to influence later priv→17 Sept 2026Does Moral Reasoning Training Help or Hurt? Red-Teaming RL-Trained Ethical Agents with Persona AttacksarXiv:2609.17552v1 Announce Type: new Abstract: Moral-reward RL can make language-model agents more cooperative, but whether that alignment survives adversarial persona pressure is unknown. Such attac→1 Oct 2026The Geometry of Harmfulness in Multi-Turn AttacksarXiv:2609.38389v1 Announce Type: cross Abstract: Large language models (LLMs) remain vulnerable to adversarial attacks that circumvent safety alignment to elicit harmful outputs. It remains unclear h→12 Oct 2026LLMail-Inject: A Dataset from a Realistic Adaptive Prompt Injection ChallengearXiv:2506.09956v2 Announce Type: replace-cross Abstract: Indirect prompt injection attacks exploit a fundamental weakness of LLMs: executing unauthorized tasks from untrusted sources. Despite the pra
CompanyMistral2 recent entries22 May 2026Beyond Benchmark Islands: Toward Representative Trustworthiness Evaluation for Agentic AIarXiv:2603.14987v2 Announce Type: replace Abstract: Agentic AI systems increasingly act through tool-augmented, multi-step workflows whose failures (unsafe tool use, unauthorised actions, social harm)→28 Sept 2026From ASR to ASP: Evaluating Prompt Attack Vulnerabilities Against Open-Source LLMsarXiv:2505.14368v3 Announce Type: replace-cross Abstract: Recent studies demonstrate that Large Language Models (LLMs) are vulnerable to attacks that generate harmful or sensitive outputs. As open-sou
CompanyxAI3 recent entries19 May 2026Multilingual jailbreaking of LLMs using low-resource languagesarXiv:2605.18239v1 Announce Type: cross Abstract: Large Language Models (LLMs) remain vulnerable to jailbreak attempts that circumvent safety guardrails. We investigate whether multi-turn conversation→28 Jul 2026Tailored untruths: How personalisation challenges LLM safeguardsarXiv:2510.12993v3 Announce Type: replace Abstract: Large Language Models (LLMs) can generate highly persuasive disinformation, yet little is known about how effectively they personalise it across lan→28 Jul 2026Gubernaut: A Deterministic Homeostatic Controller for Affect-Regulated LLM Agents, Validated Across Independent Model FamiliesarXiv:2607.24339v1 Announce Type: new Abstract: Large language model (LLM) agents inherit reactive failure modes: escalation under provocation, sycophantic drift under flattery, perseveration when stu
CompanyDeepSeek4 recent entries10 Apr 2026Break Me If You Can: Self-Jailbreaking of Aligned LLMs via Lexical Insertion PromptingarXiv:2601.02670v2 Announce Type: replace Abstract: We introduce self-jailbreaking, a threat model in which an aligned LLM guides its own compromise. Unlike most jailbreak techniques, which oft→19 May 2026Multilingual jailbreaking of LLMs using low-resource languagesarXiv:2605.18239v1 Announce Type: cross Abstract: Large Language Models (LLMs) remain vulnerable to jailbreak attempts that circumvent safety guardrails. We investigate whether multi-turn conversation→22 May 2026Beyond Benchmark Islands: Toward Representative Trustworthiness Evaluation for Agentic AIarXiv:2603.14987v2 Announce Type: replace Abstract: Agentic AI systems increasingly act through tool-augmented, multi-step workflows whose failures (unsafe tool use, unauthorised actions, social harm)→10 Jul 2026Formal Mechanisms for Market Stability in Self-Interested Agent Societies: A Marketplace Simulation StudyarXiv:2607.08652v1 Announce Type: new Abstract: Self-interested agents, left unconstrained, tend toward defection in repeated social dilemmas, causing cooperative gains from trade to collapse. This pa
CompanyNVIDIA2 recent entries29 Apr 2026The founder’s AI foundation: The top announcements for startups from Next ‘26The momentum is undeniable: the world’s fastest-growing AI startups are building with Google Cloud. Instead of stitching together fragmented point solutions, founders are building their businesses her→28 Sept 2026Why your startup needs open models alongside frontier APIsEvery week, I talk with founders who are building at an unbelievable pace. Teams are moving from inception to product-market fit faster than ever, with foundation models wired deeply into their core p