Technology & AI
MAR-12 AI detects harmful humor in memes at 80.3% accuracy
New arXiv framework uses twelve reasoning angles to separate jokes from hate in memes, hitting 80.3% humor and 75.9% hate detection on two benchmarks.
OpenAI
61 stories
Technology & AI
LLMs hold stable risk attitudes across tasks, study finds
Six LLMs tested across navigation, triage and finance tasks showed stable risk preferences, a hidden trait anyone deploying AI agents needs to understand.
Technology & AI
Text-to-video jailbreak gains 18.6% over closest rival
BSB jailbreak tricks text-to-video models like Veo and Sora by hiding harmful intent between two safe frames, beating rivals by 18.6% in attack success.
Technology & AI
OpenAI launches ChatGPT small business program
OpenAI's new ChatGPT for Small Businesses program aims to help entrepreneurs build AI skills and automate daily work, but key details are missing.
Technology & AI
AI safety thresholds: preprint proposes common frontier standard
A July 2026 preprint proposes harmonizing the capability thresholds frontier AI labs publish, which differ so much that no one can verify them.
Technology & AI
OpenAI reports failures in long-running AI models
OpenAI's July 2026 safety post details new risks in long-horizon AI models and iterative safeguards, relevant to any business deploying autonomous agents.
Technology & AI
Qubes OS: 80% of security advisories trace to upstream code
An arXiv preprint analysing 14 years of Qubes Security Bulletins finds most flaws originate in Xen and CPU components, not Qubes itself.
Technology & AI
Mycelium AI routes science context across human-agent teams
A new arXiv preprint introduces Mycelium, a shared workspace connecting researchers and AI agents, routing hypotheses to whoever can act on them.
Technology & AI
OpenAI makes case for safe teen ChatGPT access
OpenAI's July 2026 article outlines age-appropriate protections, learning tools and parental controls for teen ChatGPT users, but key questions remain.
Technology & AI
Harness Handbook maps AI agent behaviour to source code
Harness Handbook from Tencent and four universities builds a behaviour-to-code map for agent harnesses that cuts planner tokens and improves edit
Technology & AI
GPT-4o audit finds 66% of solved problems have flawed reasoning
A black-box audit swaps predicates in chain-of-thought to expose when GPT-4o reaches correct answers through reasoning that ignores its stated premises.
Technology & AI
AI agent harness evolution may not beat simple search
New arXiv paper finds automatically evolving agent scaffolding doesn't consistently outperform test-time scaling, with limited generalization to new tasks.