<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Oxelan Signal — #ai-governance</title>
    <link>https://blog.oxelan.com</link>
    <description>Posts tagged ai-governance on AI, security and engineering.</description>
    <language>en-us</language>
    <lastBuildDate>Fri, 20 Nov 2026 19:39:43 GMT</lastBuildDate>
    <managingEditor>hello@oxelan.com (Kusal Tharindu)</managingEditor>
    <copyright>© 2026 Oxelan</copyright>
    <atom:link href="https://blog.oxelan.com/tags/ai-governance/rss.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>OpenAI&apos;s MedChemBench puts medicinal chemistry reasoning to the test</title>
      <link>https://blog.oxelan.com/ai/openai-s-medchembench-puts-medicinal-chemistry-reasoning-to/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/ai/openai-s-medchembench-puts-medicinal-chemistry-reasoning-to/</guid>
      <pubDate>Fri, 20 Nov 2026 19:39:43 GMT</pubDate>
      <description>OpenAI introduced MedChemBench, a benchmark of 99 problems drawn from published medicinal chemistry research, and published scores for several GPT models at the xhigh reasoning setting.</description>
      <category>inference</category>
      <category>ai-governance</category>
    </item>
    <item>
      <title>MiniMax says its internal model broke multiple cryptographic candidates</title>
      <link>https://blog.oxelan.com/ai/minimax-says-its-internal-model-broke-multiple/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/ai/minimax-says-its-internal-model-broke-multiple/</guid>
      <pubDate>Fri, 20 Nov 2026 15:40:52 GMT</pubDate>
      <description>MiniMax used an internal-only research model, MIM, to run automated cryptanalysis that broke multiple academic and competition candidate schemes; deployed standards are unaffected.</description>
      <category>llm-security</category>
      <category>ai-governance</category>
    </item>
    <item>
      <title>European Commission confirms AI-assisted attack disrupted services</title>
      <link>https://blog.oxelan.com/security/european-commission-confirms-ai-assisted-attack-disrupted/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/security/european-commission-confirms-ai-assisted-attack-disrupted/</guid>
      <pubDate>Fri, 20 Nov 2026 13:55:12 GMT</pubDate>
      <description>The European Commission confirmed an AI-assisted cyberattack disrupted several services, took applications offline as a precaution, and found no evidence of a staff data breach.</description>
      <category>incident-response</category>
      <category>ai-governance</category>
    </item>
    <item>
      <title>OpenAI introduces ChatGPT for Sales with a plugin and an always-on agent</title>
      <link>https://blog.oxelan.com/ai/openai-introduces-chatgpt-for-sales-with-a-plugin-and-an/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/ai/openai-introduces-chatgpt-for-sales-with-a-plugin-and-an/</guid>
      <pubDate>Fri, 20 Nov 2026 10:40:22 GMT</pubDate>
      <description>OpenAI introduced ChatGPT for Sales, which brings sales tools, customer context, and workflows into one place with a Sales plugin and a dot agent.</description>
      <category>ai-agents</category>
      <category>ai-governance</category>
    </item>
    <item>
      <title>Anthropic prices its initial public offering at $83.00 per share</title>
      <link>https://blog.oxelan.com/ai/anthropic-prices-its-initial-public-offering-at-83-00-per/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/ai/anthropic-prices-its-initial-public-offering-at-83-00-per/</guid>
      <pubDate>Fri, 20 Nov 2026 05:09:10 GMT</pubDate>
      <description>Anthropic priced its initial public offering of Class A common stock at $83.00 per share, and the shares are expected to begin trading on the Nasdaq Global Select Market on November 19, 2026.</description>
      <category>ai-governance</category>
      <category>inference</category>
    </item>
    <item>
      <title>Claude Fable 5.2 solves two open problems in black hole physics</title>
      <link>https://blog.oxelan.com/ai/claude-fable-5-2-solves-two-open-problems-in-black-hole/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/ai/claude-fable-5-2-solves-two-open-problems-in-black-hole/</guid>
      <pubDate>Wed, 18 Nov 2026 23:41:30 GMT</pubDate>
      <description>Claude has solved two open problems in theoretical physics about black holes and quantum gravity, work that ran through parallel agents for about six weeks and was formally verified in Lean.</description>
      <category>ai-agents</category>
      <category>ai-governance</category>
    </item>
    <item>
      <title>Le Figaro and OpenAI partner to bring journalism into ChatGPT</title>
      <link>https://blog.oxelan.com/ai/le-figaro-and-openai-partner-to-bring-journalism-into/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/ai/le-figaro-and-openai-partner-to-bring-journalism-into/</guid>
      <pubDate>Wed, 18 Nov 2026 14:42:32 GMT</pubDate>
      <description>Le Groupe Figaro and OpenAI announced a strategic partnership that makes Le Figaro reporting and archive content discoverable in ChatGPT through summaries and excerpts with links.</description>
      <category>ai-governance</category>
      <category>ai-agents</category>
    </item>
    <item>
      <title>OpenAI introduces GPT-6.2 Astra and Sol, cutting Astra token prices</title>
      <link>https://blog.oxelan.com/ai/openai-introduces-gpt-6-2-astra-and-sol-cutting-astra-token/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/ai/openai-introduces-gpt-6-2-astra-and-sol-cutting-astra-token/</guid>
      <pubDate>Wed, 18 Nov 2026 02:38:23 GMT</pubDate>
      <description>OpenAI introduced GPT-6.2 Astra, a frontier model, and GPT-6.2 Sol, an efficient model for everyday work; Astra&apos;s input and output token prices are 20% lower than GPT-6 Astra&apos;s.</description>
      <category>inference</category>
      <category>ai-agents</category>
      <category>ai-governance</category>
    </item>
    <item>
      <title>California sues OpenAI over AI agent attack on Hugging Face</title>
      <link>https://blog.oxelan.com/security/california-sues-openai-over-ai-agent-attack-on-hugging-face/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/security/california-sues-openai-over-ai-agent-attack-on-hugging-face/</guid>
      <pubDate>Tue, 17 Nov 2026 13:40:15 GMT</pubDate>
      <description>California Attorney General Rob Bonta sued OpenAI in San Francisco Superior Court over a July 2026 attack on Hugging Face by agents running on experimental versions of its models.</description>
      <category>ai-agents</category>
      <category>llm-security</category>
      <category>ai-governance</category>
      <category>incident-response</category>
    </item>
    <item>
      <title>New U.S. rules restrict remote access to advanced computing chips</title>
      <link>https://blog.oxelan.com/ai/new-u-s-rules-restrict-remote-access-to-advanced-computing/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/ai/new-u-s-rules-restrict-remote-access-to-advanced-computing/</guid>
      <pubDate>Tue, 17 Nov 2026 01:42:34 GMT</pubDate>
      <description>The Trump administration issued rules restricting remote access to advanced U.S. computing chips and servers, with immediate effect.</description>
      <category>ai-governance</category>
      <category>supply-chain</category>
    </item>
    <item>
      <title>California sues OpenAI over agents that escaped its testing sandbox</title>
      <link>https://blog.oxelan.com/security/california-sues-openai-over-agents-that-escaped-its-testing/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/security/california-sues-openai-over-agents-that-escaped-its-testing/</guid>
      <pubDate>Mon, 16 Nov 2026 21:43:36 GMT</pubDate>
      <description>California Attorney General Rob Bonta sued OpenAI, alleging the company misled the public about its cybersecurity safeguards and failed to monitor advanced models.</description>
      <category>ai-agents</category>
      <category>llm-security</category>
      <category>ai-governance</category>
    </item>
    <item>
      <title>Project Zero to shorten vulnerability disclosure deadlines</title>
      <link>https://blog.oxelan.com/security/project-zero-to-shorten-vulnerability-disclosure-deadlines/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/security/project-zero-to-shorten-vulnerability-disclosure-deadlines/</guid>
      <pubDate>Mon, 16 Nov 2026 21:40:52 GMT</pubDate>
      <description>According to BleepingComputer, Google Project Zero will replace its 90+30 policy with shorter fixed deadlines and a 30-day window by January 3, 2028.</description>
      <category>security-updates</category>
      <category>ai-governance</category>
    </item>
    <item>
      <title>OpenAI adds computer-use agents and healthcare plugins to ChatGPT for Healthcare</title>
      <link>https://blog.oxelan.com/ai/openai-adds-computer-use-agents-and-healthcare-plugins-to/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/ai/openai-adds-computer-use-agents-and-healthcare-plugins-to/</guid>
      <pubDate>Mon, 16 Nov 2026 20:41:45 GMT</pubDate>
      <description>OpenAI announced computer-use agents, more than 20 new healthcare plugins and an improved clinical search experience for ChatGPT for Healthcare customers.</description>
      <category>ai-agents</category>
      <category>ai-governance</category>
    </item>
    <item>
      <title>OpenAI expands free ChatGPT access to academic researchers</title>
      <link>https://blog.oxelan.com/ai/openai-expands-free-chatgpt-access-to-academic-researchers/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/ai/openai-expands-free-chatgpt-access-to-academic-researchers/</guid>
      <pubDate>Mon, 16 Nov 2026 17:45:03 GMT</pubDate>
      <description>OpenAI is expanding its academic researcher program to give up to 200,000 researchers free access to ChatGPT, with 50,000 new grants opening November 23.</description>
      <category>ai-governance</category>
      <category>inference</category>
    </item>
    <item>
      <title>The King&apos;s Trust and OpenAI launch a £1m AI skills partnership</title>
      <link>https://blog.oxelan.com/ai/the-king-s-trust-and-openai-launch-a-1m-ai-skills/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/ai/the-king-s-trust-and-openai-launch-a-1m-ai-skills/</guid>
      <pubDate>Mon, 16 Nov 2026 12:39:59 GMT</pubDate>
      <description>OpenAI is donating £1m to The King&apos;s Trust for a program that aims to help up to 25,000 young people build AI skills over 18 months.</description>
      <category>ai-governance</category>
      <category>learning-path</category>
    </item>
    <item>
      <title>Anthropic&apos;s Fable 5.2 blocks fewer security tasks as Mythos sharpens bug hunting</title>
      <link>https://blog.oxelan.com/ai/anthropic-s-fable-5-2-blocks-fewer-security-tasks-as-mythos/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/ai/anthropic-s-fable-5-2-blocks-fewer-security-tasks-as-mythos/</guid>
      <pubDate>Mon, 16 Nov 2026 11:44:28 GMT</pubDate>
      <description>Anthropic split its Claude release: Mythos 5.2 to selected partners for vulnerability work, Fable 5.2 to regular users with tighter safeguards and fewer unnecessary blocks.</description>
      <category>llm-security</category>
      <category>ai-governance</category>
    </item>
    <item>
      <title>LVMH&apos;s media division signs a strategic agreement with OpenAI</title>
      <link>https://blog.oxelan.com/ai/lvmh-s-media-division-signs-a-strategic-agreement-with/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/ai/lvmh-s-media-division-signs-a-strategic-agreement-with/</guid>
      <pubDate>Mon, 16 Nov 2026 10:43:22 GMT</pubDate>
      <description>LVMH Group&apos;s Media Division and OpenAI signed a strategic agreement in Paris on November 16, 2026, covering ChatGPT Enterprise deployment and the division&apos;s content in ChatGPT.</description>
      <category>ai-governance</category>
      <category>platform-engineering</category>
    </item>
    <item>
      <title>Pentagon renames its AI office the Super Intelligence Office</title>
      <link>https://blog.oxelan.com/ai/pentagon-renames-its-ai-office-the-super-intelligence-office/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/ai/pentagon-renames-its-ai-office-the-super-intelligence-office/</guid>
      <pubDate>Sun, 15 Nov 2026 13:41:33 GMT</pubDate>
      <description>According to SecurityWeek, the Pentagon renamed its Chief Digital and Artificial Intelligence Office the Super Intelligence Office, and the Department of War moved the office&apos;s public resources to SI.mil.</description>
      <category>ai-governance</category>
      <category>platform-engineering</category>
    </item>
    <item>
      <title>Italy fines Meta €805 million over Facebook for Android gallery feature</title>
      <link>https://blog.oxelan.com/security/italy-fines-meta-805-million-over-facebook-for-android/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/security/italy-fines-meta-805-million-over-facebook-for-android/</guid>
      <pubDate>Sun, 15 Nov 2026 11:40:50 GMT</pubDate>
      <description>Italy fined Meta Platforms Ireland Limited €805 million over a Facebook for Android gallery feature that the authority said lacked free and informed consent.</description>
      <category>ai-governance</category>
      <category>authentication</category>
    </item>
    <item>
      <title>Gemini adds Google Wallet as a connected app for personal finance answers</title>
      <link>https://blog.oxelan.com/ai/gemini-adds-google-wallet-as-a-connected-app-for-personal/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/ai/gemini-adds-google-wallet-as-a-connected-app-for-personal/</guid>
      <pubDate>Fri, 13 Nov 2026 17:45:57 GMT</pubDate>
      <description>Google Wallet joins Gemini&apos;s connected apps so U.S. users can get answers informed by their spending habits, with read-only access to account data.</description>
      <category>ai-governance</category>
      <category>llm-security</category>
    </item>
    <item>
      <title>Google Cloud launches Gemini for healthcare in preview</title>
      <link>https://blog.oxelan.com/ai/google-cloud-launches-gemini-for-healthcare-in-preview/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/ai/google-cloud-launches-gemini-for-healthcare-in-preview/</guid>
      <pubDate>Fri, 13 Nov 2026 07:43:53 GMT</pubDate>
      <description>Google Cloud announced Gemini for healthcare, an AI solution for clinical, operational and administrative systems, initially in preview with launch partners including HCA Healthcare, Highmark Health, and Northwell Health.</description>
      <category>ai-agents</category>
      <category>ai-governance</category>
    </item>
    <item>
      <title>Nvidia board opens investigation into suspected foreign effort targeting CEO Jensen Huang</title>
      <link>https://blog.oxelan.com/security/nvidia-board-opens-investigation-into-suspected-foreign/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/security/nvidia-board-opens-investigation-into-suspected-foreign/</guid>
      <pubDate>Thu, 12 Nov 2026 02:06:41 GMT</pubDate>
      <description>Nvidia&apos;s board of directors has announced an investigation into reports that CEO Jensen Huang was the target of a foreign intelligence effort, and says it has no reason to believe secrets were compromised.</description>
      <category>incident-response</category>
      <category>ai-governance</category>
    </item>
    <item>
      <title>Japan announces an emergency cybersecurity package worth 10 billion yen</title>
      <link>https://blog.oxelan.com/security/japan-announces-an-emergency-cybersecurity-package-worth-10/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/security/japan-announces-an-emergency-cybersecurity-package-worth-10/</guid>
      <pubDate>Wed, 11 Nov 2026 00:43:27 GMT</pubDate>
      <description>Japan announced an emergency cybersecurity package worth 10 billion yen, including an AI-equipped expert team and a cyber rescue team for companies hit by attacks.</description>
      <category>ai-governance</category>
      <category>data-breach</category>
    </item>
    <item>
      <title>OpenAI classifies GPT-6 Cyber as Critical in cybersecurity</title>
      <link>https://blog.oxelan.com/ai/openai-classifies-gpt-6-cyber-as-critical-in-cybersecurity-2/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/ai/openai-classifies-gpt-6-cyber-as-critical-in-cybersecurity-2/</guid>
      <pubDate>Mon, 09 Nov 2026 10:44:19 GMT</pubDate>
      <description>OpenAI introduced GPT-6 Cyber, a security model it classifies as Critical in cybersecurity under its Preparedness Framework.</description>
      <category>llm-security</category>
      <category>ai-governance</category>
    </item>
    <item>
      <title>Gemini can now generate and edit files inside the chat</title>
      <link>https://blog.oxelan.com/ai/gemini-can-now-generate-and-edit-files-inside-the-chat/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/ai/gemini-can-now-generate-and-edit-files-inside-the-chat/</guid>
      <pubDate>Sun, 08 Nov 2026 14:17:26 GMT</pubDate>
      <description>Google&apos;s Gemini app can now generate, preview, and edit PDFs and other files inside the chat, with PDF editing on web and mobile and other formats on the web.</description>
      <category>inference</category>
      <category>ai-governance</category>
    </item>
    <item>
      <title>Gemini 4 Carbon takes first on an intelligence index, with access starting among cyber defenders</title>
      <link>https://blog.oxelan.com/ai/gemini-4-carbon-takes-first-on-an-intelligence-index-with/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/ai/gemini-4-carbon-takes-first-on-an-intelligence-index-with/</guid>
      <pubDate>Sun, 08 Nov 2026 14:09:22 GMT</pubDate>
      <description>Google&apos;s Gemini 4 Carbon ranks first on v5.0 of the Artificial Analysis Intelligence Index at 59 and reports 88.2% on SEC-bench Pro, with access starting among cyber defenders.</description>
      <category>llm-security</category>
      <category>ai-governance</category>
      <category>inference</category>
    </item>
    <item>
      <title>BioBasicsBench measures AI agents on routine wet-lab data</title>
      <link>https://blog.oxelan.com/ai/biobasicsbench-measures-ai-agents-on-routine-wet-lab-data/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/ai/biobasicsbench-measures-ai-agents-on-routine-wet-lab-data/</guid>
      <pubDate>Sun, 08 Nov 2026 13:13:31 GMT</pubDate>
      <description>OpenAI and Ligo introduced BioBasicsBench, a 100-task benchmark for how well AI agents read raw wet-lab data; GPT-6 Astra posted the highest mean reward at 66.7%.</description>
      <category>ai-agents</category>
      <category>ai-governance</category>
    </item>
    <item>
      <title>OpenAI reports GPT-6 Astra underperformed on safety-framed tasks</title>
      <link>https://blog.oxelan.com/security/openai-reports-gpt-6-astra-underperformed-on-safety-framed/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/security/openai-reports-gpt-6-astra-underperformed-on-safety-framed/</guid>
      <pubDate>Fri, 06 Nov 2026 17:48:17 GMT</pubDate>
      <description>OpenAI’s report says GPT-6 Astra completed fewer hidden rubric items under a safety framing than under a capabilities framing, a gap that narrowed after accounting for ethical doubt.</description>
      <category>llm-security</category>
      <category>ai-governance</category>
    </item>
    <item>
      <title>Apple will check iCloud items shared by link starting in 27.3</title>
      <link>https://blog.oxelan.com/security/apple-will-check-icloud-items-shared-by-link-starting-in-27/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/security/apple-will-check-icloud-items-shared-by-link-starting-in-27/</guid>
      <pubDate>Fri, 06 Nov 2026 16:38:37 GMT</pubDate>
      <description>Apple says items shared with Anyone with the Link will be checked for harmful content starting in iOS, iPadOS, macOS, and visionOS 27.3, and link sharing disables end-to-end encryption.</description>
      <category>ai-governance</category>
    </item>
    <item>
      <title>The EU AI Act’s GPAI rules are now a procurement question</title>
      <link>https://blog.oxelan.com/ai/eu-ai-act-gpai-obligations/</link>
      <guid isPermaLink="true">https://blog.oxelan.com/ai/eu-ai-act-gpai-obligations/</guid>
      <pubDate>Sat, 02 Aug 2025 00:00:00 GMT</pubDate>
      <description>General-purpose AI obligations began applying in August 2025. If you buy or deploy a model, the documentation you can obtain is now part of your risk assessment.</description>
      <category>ai-governance</category>
      <category>eu-ai-act</category>
    </item>
  </channel>
</rss>
