<?xml version="1.0" encoding="UTF-8"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Safety — Lab Ledger Desk</title><link>https://labledgerdesk.pages.dev/topic/safety/</link><description>Official AI-lab briefs tagged Safety from allow-listed sources.</description><atom:link href="https://labledgerdesk.pages.dev/topic/safety/rss.xml" rel="self" type="application/rss+xml"/><item><title>Responsible AI for agentic AI: A phased roadmap for platform and application teams</title><link>https://labledgerdesk.pages.dev/b/2026/11/17/responsible-ai-for-agentic-ai-a-phased-roadmap-for-platform-and/</link><guid>https://labledgerdesk.pages.dev/b/2026/11/17/responsible-ai-for-agentic-ai-a-phased-roadmap-for-platform-and/</guid><category>Note</category><description>Enterprises adopting agentic AI at scale need platforms that embed responsible AI from the start. This post gives platform and application teams a…</description><pubDate>Tue, 17 Nov 2026 16:56:31 GMT</pubDate></item><item><title>Introducing OpenAI for Chile</title><link>https://labledgerdesk.pages.dev/b/2026/11/17/introducing-openai-for-chile/</link><guid>https://labledgerdesk.pages.dev/b/2026/11/17/introducing-openai-for-chile/</guid><category>Launch</category><description>OpenAI and the Government of Chile are partnering to support responsible AI adoption, build public-sector skills and help local startups develop AI…</description><pubDate>Tue, 17 Nov 2026 00:00:00 GMT</pubDate></item><item><title>California sues OpenAI over rogue AI hacking incidents</title><link>https://labledgerdesk.pages.dev/b/2026/11/16/california-sues-openai-over-rogue-ai-hacking-incidents/</link><guid>https://labledgerdesk.pages.dev/b/2026/11/16/california-sues-openai-over-rogue-ai-hacking-incidents/</guid><category>Note</category><description>California is suing OpenAI after its AI agents broke out of their testing environment and hacked other companies, including Hugging Face. In a lawsuit filed…</description><pubDate>Mon, 16 Nov 2026 20:46:01 GMT</pubDate></item><item><title>Building trustworthy AI by fencing it in, not predicting it</title><link>https://labledgerdesk.pages.dev/b/2026/11/16/building-trustworthy-ai-by-fencing-it-in-not-predicting-it/</link><guid>https://labledgerdesk.pages.dev/b/2026/11/16/building-trustworthy-ai-by-fencing-it-in-not-predicting-it/</guid><category>Note</category><description>In October this year, OpenAI disclosed that two of its AI agents had switched off the systems built to monitor them. One muted its own safety alerts after…</description><pubDate>Mon, 16 Nov 2026 16:04:28 GMT</pubDate></item><item><title>The OpenAI board member who fired Sam Altman says we still can’t trust AI CEOs</title><link>https://labledgerdesk.pages.dev/b/2026/11/14/the-openai-board-member-who-fired-sam-altman-says-we-still-can-t-trust/</link><guid>https://labledgerdesk.pages.dev/b/2026/11/14/the-openai-board-member-who-fired-sam-altman-says-we-still-can-t-trust/</guid><category>Note</category><description>Today, I’m talking with Helen Toner, the executive director of the Center for Security and Emerging Technology at Georgetown University. This episode is…</description><pubDate>Sat, 14 Nov 2026 15:00:00 GMT</pubDate></item><item><title>Supporting research on AI and children’s safety</title><link>https://labledgerdesk.pages.dev/b/2026/11/12/supporting-research-on-ai-and-children-s-safety/</link><guid>https://labledgerdesk.pages.dev/b/2026/11/12/supporting-research-on-ai-and-children-s-safety/</guid><category>Research</category><description>OpenAI commits $2 million to research on AI and children’s safety. The work supports shared tools to measure risks and opportunities for young people.</description><pubDate>Thu, 12 Nov 2026 00:00:00 GMT</pubDate></item><item><title>Theoretical Limits of Language Model Alignment</title><link>https://labledgerdesk.pages.dev/b/2026/11/12/theoretical-limits-of-language-model-alignment/</link><guid>https://labledgerdesk.pages.dev/b/2026/11/12/theoretical-limits-of-language-model-alignment/</guid><category>Note</category><description>Language model (LM) alignment improves model outputs to reflect human preferences while preserving the capabilities of the base model. The most common…</description><pubDate>Thu, 12 Nov 2026 00:00:00 GMT</pubDate></item><item><title>OpenAI, Anthropic face new oversight as New York’s AI safety law takes effect</title><link>https://labledgerdesk.pages.dev/b/2026/11/10/openai-anthropic-face-new-oversight-as-new-york-s-ai-safety-law-takes/</link><guid>https://labledgerdesk.pages.dev/b/2026/11/10/openai-anthropic-face-new-oversight-as-new-york-s-ai-safety-law-takes/</guid><category>Note</category><description>New York is preparing to enforce its sweeping AI safety law, requiring major developers like OpenAI and Anthropic to register, disclose safety practices…</description><pubDate>Tue, 10 Nov 2026 20:52:13 GMT</pubDate></item><item><title>The AI Safety Debate Has Lost the Plot. We Asked 10 Experts What the Real Nightmares Are</title><link>https://labledgerdesk.pages.dev/b/2026/11/10/the-ai-safety-debate-has-lost-the-plot-we-asked-10-experts-what-the/</link><guid>https://labledgerdesk.pages.dev/b/2026/11/10/the-ai-safety-debate-has-lost-the-plot-we-asked-10-experts-what-the/</guid><category>Note</category><description>The risk of nuclear catastrophe is real, as is the fallout of the China–US domination game (hi there, spies). But no need to panic about Skynet.</description><pubDate>Tue, 10 Nov 2026 11:00:00 GMT</pubDate></item><item><title>Gemini Robotics 2.5 accelerates real-world general dexterity</title><link>https://labledgerdesk.pages.dev/b/2026/11/09/gemini-robotics-2-5-accelerates-real-world-general-dexterity/</link><guid>https://labledgerdesk.pages.dev/b/2026/11/09/gemini-robotics-2-5-accelerates-real-world-general-dexterity/</guid><category>Note</category><description>Our most capable vision-language-action model enhances robot dexterity, reliability, and safety</description><pubDate>Mon, 09 Nov 2026 19:51:47 GMT</pubDate></item><item><title>Trump’s 60 Minutes interview touched on AI safety, threat to arrest an AI exec</title><link>https://labledgerdesk.pages.dev/b/2026/11/09/trump-s-60-minutes-interview-touched-on-ai-safety-threat-to-arrest-an/</link><guid>https://labledgerdesk.pages.dev/b/2026/11/09/trump-s-60-minutes-interview-touched-on-ai-safety-threat-to-arrest-an/</guid><category>Note</category><description>Asked about AI executives' warnings about their models' safety on 60 Minutes, Trump said, &quot;I think that the risk is not nearly as real as it- as they talk…</description><pubDate>Mon, 09 Nov 2026 15:50:38 GMT</pubDate></item><item><title>Introducing Control Room: Human control over AI workloads</title><link>https://labledgerdesk.pages.dev/b/2026/11/09/introducing-control-room-human-control-over-ai-workloads/</link><guid>https://labledgerdesk.pages.dev/b/2026/11/09/introducing-control-room-human-control-over-ai-workloads/</guid><category>Launch</category><description>Learn how OpenAI’s Control Room tracks model workloads across our research infrastructure and can rapidly stop them when safety or security concerns arise.</description><pubDate>Mon, 09 Nov 2026 13:00:00 GMT</pubDate></item><item><title>Control Room: A central system to track and stop model workloads across our research infrastructure</title><link>https://labledgerdesk.pages.dev/b/2026/11/09/control-room-a-central-system-to-track-and-stop-model-workloads-across/</link><guid>https://labledgerdesk.pages.dev/b/2026/11/09/control-room-a-central-system-to-track-and-stop-model-workloads-across/</guid><category>Research</category><description>Learn how OpenAI’s Control Room tracks model workloads across our research infrastructure and can rapidly stop them when safety or security concerns arise.</description><pubDate>Mon, 09 Nov 2026 13:00:00 GMT</pubDate></item><item><title>Why a new report on Anthropic automating AI development is alarming researchers</title><link>https://labledgerdesk.pages.dev/b/2026/11/07/why-a-new-report-on-anthropic-automating-ai-development-is-alarming/</link><guid>https://labledgerdesk.pages.dev/b/2026/11/07/why-a-new-report-on-anthropic-automating-ai-development-is-alarming/</guid><category>Research</category><description>Anthropic is using AI to build more powerful AI. And it appears to be working remarkably well, with the company's models improving roughly twice as quickly…</description><pubDate>Sat, 07 Nov 2026 12:30:00 GMT</pubDate></item><item><title>Taxbit Assist: A grounded, role-aware AI agent on Amazon Bedrock AgentCore</title><link>https://labledgerdesk.pages.dev/b/2026/11/06/taxbit-assist-a-grounded-role-aware-ai-agent-on-amazon-bedrock-agentcore/</link><guid>https://labledgerdesk.pages.dev/b/2026/11/06/taxbit-assist-a-grounded-role-aware-ai-agent-on-amazon-bedrock-agentcore/</guid><category>Note</category><description>Taxbit built Taxbit Assist, a cited, permission-aware AI assistant embedded in its enterprise dashboard, on Amazon Bedrock AgentCore. …</description><pubDate>Fri, 06 Nov 2026 16:49:42 GMT</pubDate></item><item><title>Setting a new standard for responsible AI infrastructure</title><link>https://labledgerdesk.pages.dev/b/2026/11/05/setting-a-new-standard-for-responsible-ai-infrastructure/</link><guid>https://labledgerdesk.pages.dev/b/2026/11/05/setting-a-new-standard-for-responsible-ai-infrastructure/</guid><category>Launch</category><description>OpenAI introduces its National Community Compact and Compute Campus Policy Blueprint to protect local resources, create jobs, and guide responsible AI…</description><pubDate>Thu, 05 Nov 2026 02:00:00 GMT</pubDate></item><item><title>How “reverse game theory” could keep AI agents in check</title><link>https://labledgerdesk.pages.dev/b/2026/11/03/how-reverse-game-theory-could-keep-ai-agents-in-check/</link><guid>https://labledgerdesk.pages.dev/b/2026/11/03/how-reverse-game-theory-could-keep-ai-agents-in-check/</guid><category>Note</category><description>A new approach known as “reverse game theory” aims to control powerful AI models by designing systems that incentivize them to stick to the rules. As AI…</description><pubDate>Tue, 03 Nov 2026 10:43:41 GMT</pubDate></item></channel></rss>