<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0" xml:base="https://dailyaimail.news"
  xmlns:dc="http://purl.org/dc/elements/1.1/"
  xmlns:atom="http://www.w3.org/2005/Atom"
  xmlns:media="http://search.yahoo.com/mrss/">
  <channel>
    <title><![CDATA[AI Safety News - Daily AI Mail]]></title>
    <link>https://dailyaimail.news/topics/ai-safety</link>
    <description><![CDATA[Latest AI Safety news, analysis, and updates from Daily AI Mail.]]></description>
    <language>en</language>
    <atom:link href="https://dailyaimail.news/topics/ai-safety/rss.xml" rel="self" type="application/rss+xml" />
    <item>
      <title><![CDATA[Florida Sues OpenAI and Sam Altman Over ChatGPT's Alleged Role in Violent Incidents]]></title>
      <link><![CDATA[https://dailyaimail.news/news/florida-openai-lawsuit-chatgpt-violence]]></link>
      <description><![CDATA[Florida's attorney general has filed what he calls the first state-led lawsuit against OpenAI and CEO Sam Altman over ChatGPT's alleged links to violent incidents. The case pushes AI safety litigation into a new phase by targeting a frontier lab through state action rather than only private civil claims.]]></description>
      <pubDate>Mon, 01 Jun 2026 23:06:00 GMT</pubDate>
      <dc:creator><![CDATA[Oliver Randall]]></dc:creator>
      <category><![CDATA[AI Safety]]></category>
      <guid isPermaLink="false"><![CDATA[https://dailyaimail.news/news/florida-openai-lawsuit-chatgpt-violence]]></guid>
      <media:thumbnail url="https://dailyaimail.news/_astro/florida-sues-openai-sam-altman.G2_VMKaq.webp" />
      <media:content url="https://dailyaimail.news/_astro/florida-sues-openai-sam-altman.G2_VMKaq.webp" medium="image">
        <media:title type="html"><![CDATA[florida-sues-openai-sam-altman.G2_VMKaq.webp]]></media:title>
      </media:content>
    </item>
    <item>
      <title><![CDATA[OpenAI Opens GPT-5.5 Bio Bug Bounty to Test Frontier AI Safety Controls]]></title>
      <link><![CDATA[https://dailyaimail.news/news/openai-gpt-5-5-bio-bug-bounty]]></link>
      <description><![CDATA[OpenAI is inviting vetted red-teamers, security researchers, and biosecurity specialists to test whether GPT-5.5 can be universally jailbroken on five biology safety questions. The private bounty offers $25,000 for the first true universal jailbreak.]]></description>
      <pubDate>Fri, 24 Apr 2026 19:39:00 GMT</pubDate>
      <dc:creator><![CDATA[Daily AI Mail Editorial Staff]]></dc:creator>
      <category><![CDATA[AI Safety]]></category>
      <guid isPermaLink="false"><![CDATA[https://dailyaimail.news/news/openai-gpt-5-5-bio-bug-bounty]]></guid>
      <media:thumbnail url="https://dailyaimail.news/_astro/gpt-5-5-bio-bug-bounty.BqHwnQiR.png" />
      <media:content url="https://dailyaimail.news/_astro/gpt-5-5-bio-bug-bounty.BqHwnQiR.png" medium="image">
        <media:title type="html"><![CDATA[gpt-5-5-bio-bug-bounty.BqHwnQiR.png]]></media:title>
      </media:content>
    </item>
    <item>
      <title><![CDATA[Anthropic Unveils Project Glasswing to Put Frontier AI on Cyber Defense Duty]]></title>
      <link><![CDATA[https://dailyaimail.news/news/anthropic-project-glasswing-cyber-defense]]></link>
      <description><![CDATA[Anthropic has launched Project Glasswing, a new initiative built around Claude Mythos Preview to help secure critical software before advanced AI systems make cyberattacks easier to scale. The company is framing it as a defense-first response to rapidly improving AI vulnerability research.]]></description>
      <pubDate>Tue, 07 Apr 2026 22:55:00 GMT</pubDate>
      <dc:creator><![CDATA[Daily AI Mail Editorial Staff]]></dc:creator>
      <category><![CDATA[AI Safety]]></category>
      <guid isPermaLink="false"><![CDATA[https://dailyaimail.news/news/anthropic-project-glasswing-cyber-defense]]></guid>
      <media:thumbnail url="https://dailyaimail.news/_astro/anthropic-project-glasswing-cyber-defense.BGul2kCP.png" />
      <media:content url="https://dailyaimail.news/_astro/anthropic-project-glasswing-cyber-defense.BGul2kCP.png" medium="image">
        <media:title type="html"><![CDATA[anthropic-project-glasswing-cyber-defense.BGul2kCP.png]]></media:title>
      </media:content>
    </item>
    <item>
      <title><![CDATA[OpenAI Launches Safety Bug Bounty Program to Reward Researchers Who Find AI Abuse Risks]]></title>
      <link><![CDATA[https://dailyaimail.news/news/openai-safety-bug-bounty-program-launch]]></link>
      <description><![CDATA[OpenAI is opening a public Safety Bug Bounty program targeting AI-specific misuse scenarios — from agentic prompt injection to platform integrity bypasses — that fall outside traditional security vulnerability scopes.]]></description>
      <pubDate>Tue, 07 Apr 2026 22:50:27 GMT</pubDate>
      <dc:creator><![CDATA[Daily AI Mail Editorial Staff]]></dc:creator>
      <category><![CDATA[AI Safety]]></category>
      <guid isPermaLink="false"><![CDATA[https://dailyaimail.news/news/openai-safety-bug-bounty-program-launch]]></guid>
      <media:thumbnail url="https://dailyaimail.news/_astro/openai-safety-bug-bounty-program-launch.BLpR8BjO.png" />
      <media:content url="https://dailyaimail.news/_astro/openai-safety-bug-bounty-program-launch.BLpR8BjO.png" medium="image">
        <media:title type="html"><![CDATA[openai-safety-bug-bounty-program-launch.BLpR8BjO.png]]></media:title>
      </media:content>
    </item>
    <item>
      <title><![CDATA[Claude Mythos Leaked: Anthropic's Most Powerful Model Yet Poses 'Unprecedented Cybersecurity Risk']]></title>
      <link><![CDATA[https://dailyaimail.news/news/claude-mythos-capybara-leak-anthropic-frontier-model]]></link>
      <description><![CDATA[A CMS misconfiguration exposed nearly 3,000 internal Anthropic assets, including a draft blog post describing Claude Mythos — a new model tier above Opus that the company itself warns is 'far ahead of any other AI model in cyber capabilities.' Anthropic has confirmed the model exists.]]></description>
      <pubDate>Wed, 01 Apr 2026 12:00:00 GMT</pubDate>
      <dc:creator><![CDATA[Daily AI Mail Editorial Staff]]></dc:creator>
      <category><![CDATA[AI Safety]]></category>
      <guid isPermaLink="false"><![CDATA[https://dailyaimail.news/news/claude-mythos-capybara-leak-anthropic-frontier-model]]></guid>
      <media:thumbnail url="https://dailyaimail.news/_astro/claude-mythos-capybara-leak-anthropic-frontier-model.DphZ80--.png" />
      <media:content url="https://dailyaimail.news/_astro/claude-mythos-capybara-leak-anthropic-frontier-model.DphZ80--.png" medium="image">
        <media:title type="html"><![CDATA[claude-mythos-capybara-leak-anthropic-frontier-model.DphZ80--.png]]></media:title>
      </media:content>
    </item>
    <item>
      <title><![CDATA[Stanford Study: AI Chatbots Are Making People More Self-Centered — and Tech Companies Have Every Reason to Keep It That Way]]></title>
      <link><![CDATA[https://dailyaimail.news/news/stanford-ai-sycophancy-study-self-centered]]></link>
      <description><![CDATA[A peer-reviewed Stanford paper published in Science finds that AI sycophancy doesn't just flatter users — it makes them less likely to apologize, more morally rigid, and increasingly dependent on machine validation over human feedback. The researchers say it's a safety issue that needs regulation.]]></description>
      <pubDate>Sun, 29 Mar 2026 19:47:00 GMT</pubDate>
      <dc:creator><![CDATA[Daily AI Mail Editorial Staff]]></dc:creator>
      <category><![CDATA[AI Safety]]></category>
      <guid isPermaLink="false"><![CDATA[https://dailyaimail.news/news/stanford-ai-sycophancy-study-self-centered]]></guid>
      <media:thumbnail url="https://dailyaimail.news/_astro/stanford-ai-sycophancy-study-self-centered.BhdeQT6b.png" />
      <media:content url="https://dailyaimail.news/_astro/stanford-ai-sycophancy-study-self-centered.BhdeQT6b.png" medium="image">
        <media:title type="html"><![CDATA[stanford-ai-sycophancy-study-self-centered.BhdeQT6b.png]]></media:title>
      </media:content>
    </item>
  </channel>
</rss>