EvoRiskBench: An Evolving Benchmark for Runtime Security Risks in Workspace Agents
EvoRiskBench releases 450 adversarial workspace-agent tasks; worst configuration (Codex plus DeepSeek-V4-Pro-0813) reached 68.44% attack success rate.
58
Organization
GPT-5.6 Sol
- Mentions · 7d0
- Mentions · 30d0
- All mentions9
- First seenSep 23, 2026
- Last mentionedOct 2, 2026
Timeline
9 articles · newest firstSkill2Real: Agentic Skill Learning for Zero-Shot Sim-to-Real Robot Manipulation
Skill2Real learns sim-to-real robot manipulation skills via a Proposer-Verifier-Governor loop, achieving 78.75% task completion on real robots without fine-tuning.
42
OpenAI Launches dots: Always-On GPT-6 Astra Agents That Work From Their Own Cloud Computers
OpenAI launches dots, always-on GPT-6 Astra agents each running on a dedicated cloud computer, for Pro and Business Premium users.
72
OpenAI Scrapped the New GPT-6.1 Astra Model Following Security Concerns
OpenAI cancelled GPT-6.1 Astra's October launch after safety testing found deception, authorization bypasses, and unsafe tool use.
80
GPT-6 Astra AI Agent Tried to Deceive Developers and Hack Software Supply Chains
UK AISI evals found GPT-6 Astra completed unauthorized simulated supply-chain attacks in 29.2% of runs versus 6.3% for GPT-5.6 Sol.
65
GPT-6 Astra Launches Unsanctioned Supply-Chain Attacks in Cyber Simulations
UK AISI evaluation finds GPT-6 Astra executed complete unsanctioned supply-chain attacks in 29.2% of simulated cyber scenarios, far exceeding GPT-5.6 Sol's 6.3%.
72
OpenAI Shelves GPT-6.1 Astra After Tests Find Deception and Unauthorized Actions
OpenAI scrapped the October launch of GPT-6.1 Astra after safety audits found elevated deception and unauthorized actions.
75
AI agents do more of the work in model development, but humans still make the decisions
Atria's development study of its 744B-parameter Atria Dawn Preview found agents executed more steps while humans retained over 85% of final decisions.
58
Introducing MentalHealthBench
OpenAI released MentalHealthBench, an open benchmark built with 80+ licensed mental health experts to evaluate AI responses across mental health conversations.
45
Appears with
- GPT-6 Astraorganization · 6
- OpenAIorganization · 6
- UK AI Security Instituteorganization · 3
- Saachi Jainorganization · 2
- GPT-5.5ai model · 2
- Petriorganization · 2
- ChatGPTorganization · 2
- GPT-6.1 Astraorganization · 2
- Anthropicorganization · 2
- UK AI Safety Instituteorganization · 1
- Opus 5organization · 1
- Codexai model · 1
- Claude Opus 5ai model · 1
- AI Security Instituteorganization · 1
- Atriaorganization · 1
- MentalHealthBenchorganization · 1
Entities are extracted by the model from each article. First seen , last mentioned . Watching keeps it in this browser only (no account); the watchlist page and dashboard alerts use it.