<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>Zero Days — AI Governance Intelligence</title>
    <link>https://zero-days.ai</link>
    <atom:link href="https://zero-days.ai/feed.xml" rel="self" type="application/rss+xml"/>
    <description>AI safety intelligence. Before the exploit. Daily signals on EU AI Act, DSA, KOSA, FTC, agentic AI, synthetic media, and platform Trust &amp; Safety.</description>
    <language>en-us</language>
    <copyright>Zero Days 2026</copyright>
    <managingEditor>hello@zero-days.ai (Zero Days)</managingEditor>
    <lastBuildDate>Sun, 15 Mar 2026 06:10:00 GMT</lastBuildDate>
    <ttl>360</ttl>

    <item>
      <title>EU Parliament ratifies first legally binding international AI governance treaty</title>
      <link>https://www.febis.org/2026/03/12/eu-endorses-first-international-treaty-on-ai-governance-what-it-means-for-the-future-of-trusted-ai/</link>
      <guid isPermaLink="true">https://www.febis.org/2026/03/12/eu-endorses-first-international-treaty-on-ai-governance-what-it-means-for-the-future-of-trusted-ai/</guid>
      <dc:creator>Zero Days</dc:creator>
      <category>EU AI Act</category>
      <pubDate>Wed, 11 Mar 2026 00:00:00 GMT</pubDate>
      <description><![CDATA[The Council of Europe's Framework Convention on AI becomes the first international legally binding treaty on AI governance, establishing risk management obligations, human rights safeguards, and cross-border accountability requirements. <a href="https://zero-days.ai/#news-library">Read all signals →</a>]]></description>
    </item>

    <item>
      <title>FTC issues policy statement on AI Act scope and federal preemption of state AI laws</title>
      <link>https://www.mintz.com/insights-center/viewpoints/54731/2026-02-13-emerging-federal-ai-strategy-ftc-sets-aside-rytr</link>
      <guid isPermaLink="true">https://www.mintz.com/insights-center/viewpoints/54731/2026-02-13-emerging-federal-ai-strategy-ftc-sets-aside-rytr</guid>
      <dc:creator>Zero Days</dc:creator>
      <category>FTC / US Regulation</category>
      <pubDate>Wed, 11 Mar 2026 00:00:00 GMT</pubDate>
      <description><![CDATA[FTC's statement attempts to clarify where federal AI governance authority begins and state-level laws end. Preemption claims are contested — comment period deadline March 11, 2026. <a href="https://zero-days.ai/#news-library">Read all signals →</a>]]></description>
    </item>

    <item>
      <title>DSA: EC investigation into X/Grok recommender systems escalates to adversarial posture</title>
      <link>https://digital-strategy.ec.europa.eu/en/policies/dsa-enforcement</link>
      <guid isPermaLink="true">https://digital-strategy.ec.europa.eu/en/policies/dsa-enforcement</guid>
      <dc:creator>Zero Days</dc:creator>
      <category>DSA</category>
      <pubDate>Mon, 09 Mar 2026 00:00:00 GMT</pubDate>
      <description><![CDATA[EC enforcement team formally shifts to 'more adversarial' mode on X investigation. Focus on Grok integration into recommender systems and whether this constitutes an undisclosed systemic risk under DSA Article 34. <a href="https://zero-days.ai/#news-library">Read all signals →</a>]]></description>
    </item>

    <item>
      <title>EU Digital Omnibus proposal could delay AI Act Annex III high-risk obligations to Dec 2027</title>
      <link>https://www.legalnodes.com/article/eu-ai-act-2026-updates-compliance-requirements-and-business-risks</link>
      <guid isPermaLink="true">https://www.legalnodes.com/article/eu-ai-act-2026-updates-compliance-requirements-and-business-risks</guid>
      <dc:creator>Zero Days</dc:creator>
      <category>EU AI Act</category>
      <pubDate>Sun, 08 Mar 2026 00:00:00 GMT</pubDate>
      <description><![CDATA[Commission's Digital Omnibus simplification package proposes 12-month extension for Annex III high-risk AI system compliance — affecting HR tools, credit scoring, and law enforcement AI. Implementation timeline in flux. <a href="https://zero-days.ai/#news-library">Read all signals →</a>]]></description>
    </item>

    <item>
      <title>House committee passes KIDS Act 28-24 — algorithm opt-out and annual FTC audits included</title>
      <link>https://rollcall.com/2026/03/06/kids-online-safety-bills-move-forward-from-senate-house-panel/</link>
      <guid isPermaLink="true">https://rollcall.com/2026/03/06/kids-online-safety-bills-move-forward-from-senate-house-panel/</guid>
      <dc:creator>Zero Days</dc:creator>
      <category>KOSA / Child Safety</category>
      <pubDate>Fri, 06 Mar 2026 00:00:00 GMT</pubDate>
      <description><![CDATA[The updated KIDS Online Safety Act clears committee with bipartisan support, adding mandatory algorithm opt-out for minors and annual FTC platform audits. Senate companion bill advancing in parallel. <a href="https://zero-days.ai/#news-library">Read all signals →</a>]]></description>
    </item>

    <item>
      <title>OpenAI updates agentic AI usage policies — new restrictions on tool chaining and memory</title>
      <link>https://openai.com/policies/usage-policies</link>
      <guid isPermaLink="true">https://openai.com/policies/usage-policies</guid>
      <dc:creator>Zero Days</dc:creator>
      <category>Agentic AI</category>
      <pubDate>Thu, 05 Mar 2026 00:00:00 GMT</pubDate>
      <description><![CDATA[Updated policies require operators deploying agentic systems to implement human oversight checkpoints for irreversible actions. Memory persistence and multi-agent orchestration now subject to additional disclosure requirements. <a href="https://zero-days.ai/#news-library">Read all signals →</a>]]></description>
    </item>

    <item>
      <title>UK Ofcom publishes final Online Safety Act codes of practice for illegal content</title>
      <link>https://www.ofcom.org.uk/online-safety/information-for-industry/the-register-of-regulated-services</link>
      <guid isPermaLink="true">https://www.ofcom.org.uk/online-safety/information-for-industry/the-register-of-regulated-services</guid>
      <dc:creator>Zero Days</dc:creator>
      <category>Platform Incidents</category>
      <pubDate>Wed, 04 Mar 2026 00:00:00 GMT</pubDate>
      <description><![CDATA[Ofcom's binding codes cover CSAM, terrorism, fraud, and drugs content. Platforms with UK users must comply by September 2026. User-to-user and search services both in scope. Fines up to £18M or 10% of global turnover. <a href="https://zero-days.ai/#news-library">Read all signals →</a>]]></description>
    </item>

    <item>
      <title>Meta announces mandatory AI-generated content labels for all political advertising globally</title>
      <link>https://about.fb.com/news/2026/03/ai-content-labeling-political-ads/</link>
      <guid isPermaLink="true">https://about.fb.com/news/2026/03/ai-content-labeling-political-ads/</guid>
      <dc:creator>Zero Days</dc:creator>
      <category>Synthetic Media</category>
      <pubDate>Mon, 02 Mar 2026 00:00:00 GMT</pubDate>
      <description><![CDATA[Meta's expanded synthetic media policy requires all political advertisers to disclose AI-generated imagery, audio, and video. Policy applies to all elections globally from Q2 2026. Enforcement via automated detection + human review. <a href="https://zero-days.ai/#news-library">Read all signals →</a>]]></description>
    </item>

    <item>
      <title>DSA Transparency Database surpasses 1 billion moderation decisions — EC publishes analysis</title>
      <link>https://transparency.dsa.ec.europa.eu/</link>
      <guid isPermaLink="true">https://transparency.dsa.ec.europa.eu/</guid>
      <dc:creator>Zero Days</dc:creator>
      <category>DSA</category>
      <pubDate>Sun, 01 Mar 2026 00:00:00 GMT</pubDate>
      <description><![CDATA[The EC's public database hits 1B decisions milestone. Analysis shows 73% of decisions lack detailed ground classification — regulators flagging this as a transparency gap. Expect follow-up enforcement guidance Q2. <a href="https://zero-days.ai/#news-library">Read all signals →</a>]]></description>
    </item>

    <item>
      <title>NIST releases AI Risk Management Framework 1.1 with generative AI addendum</title>
      <link>https://www.nist.gov/artificial-intelligence/ai-risk-management-framework</link>
      <guid isPermaLink="true">https://www.nist.gov/artificial-intelligence/ai-risk-management-framework</guid>
      <dc:creator>Zero Days</dc:creator>
      <category>AI Safety</category>
      <pubDate>Sat, 28 Feb 2026 00:00:00 GMT</pubDate>
      <description><![CDATA[RMF 1.1 adds a GenAI profile covering hallucination risk, data provenance, and adversarial prompt injection. Maps to EU AI Act Article 9 risk management requirements — operationally useful for dual-compliance programs. <a href="https://zero-days.ai/#news-library">Read all signals →</a>]]></description>
    </item>

    <item>
      <title>FTC signals no COPPA enforcement against platforms using robust age-verification systems</title>
      <link>https://www.dwt.com/blogs/privacy--security-law-blog/2026/03/trust-issues-march-2026</link>
      <guid isPermaLink="true">https://www.dwt.com/blogs/privacy--security-law-blog/2026/03/trust-issues-march-2026</guid>
      <dc:creator>Zero Days</dc:creator>
      <category>FTC / US Regulation</category>
      <pubDate>Wed, 25 Feb 2026 00:00:00 GMT</pubDate>
      <description><![CDATA[FTC policy guidance clarifies that platforms deploying verified age assurance systems will be treated as acting in good faith under COPPA. Practical enforcement signal for operators building age verification infrastructure. <a href="https://zero-days.ai/#news-library">Read all signals →</a>]]></description>
    </item>

    <item>
      <title>California AB 2655 (DEFIANCE Act) signed — AI deepfake victims can now sue platforms</title>
      <link>https://leginfo.legislature.ca.gov/</link>
      <guid isPermaLink="true">https://leginfo.legislature.ca.gov/</guid>
      <dc:creator>Zero Days</dc:creator>
      <category>Synthetic Media</category>
      <pubDate>Fri, 20 Feb 2026 00:00:00 GMT</pubDate>
      <description><![CDATA[Amended DEFIANCE Act creates civil liability for platforms that host non-consensual AI-generated intimate images. Platforms have 24-hour removal obligation and must maintain reporting mechanisms. Effective July 1, 2026. <a href="https://zero-days.ai/#news-library">Read all signals →</a>]]></description>
    </item>

    <item>
      <title>EU AI Office publishes first GPAI model evaluation methodology for systemic risk</title>
      <link>https://digital-strategy.ec.europa.eu/en/policies/ai-office</link>
      <guid isPermaLink="true">https://digital-strategy.ec.europa.eu/en/policies/ai-office</guid>
      <dc:creator>Zero Days</dc:creator>
      <category>EU AI Act</category>
      <pubDate>Wed, 18 Feb 2026 00:00:00 GMT</pubDate>
      <description><![CDATA[The methodology covers red-teaming protocols, capability thresholds for 10^25 FLOP models, and incident reporting timelines. Frontier model providers must conduct evaluations before EU deployment. Comment period open until April 15. <a href="https://zero-days.ai/#news-library">Read all signals →</a>]]></description>
    </item>

    <item>
      <title>Amazon, eBay, and Etsy face DSA joint enforcement investigation on fake reviews</title>
      <link>https://digital-strategy.ec.europa.eu/en/policies/dsa-enforcement</link>
      <guid isPermaLink="true">https://digital-strategy.ec.europa.eu/en/policies/dsa-enforcement</guid>
      <dc:creator>Zero Days</dc:creator>
      <category>DSA</category>
      <pubDate>Sat, 14 Feb 2026 00:00:00 GMT</pubDate>
      <description><![CDATA[EC opens coordinated investigation into three major marketplaces over alleged failure to adequately surface and remove fake user reviews — a 'dark pattern' under DSA Article 25. Formal decisions expected Q3 2026. <a href="https://zero-days.ai/#news-library">Read all signals →</a>]]></description>
    </item>

    <item>
      <title>Anthropic publishes Constitutional AI v2 — new honesty and oversight mechanisms disclosed</title>
      <link>https://www.anthropic.com/research</link>
      <guid isPermaLink="true">https://www.anthropic.com/research</guid>
      <dc:creator>Zero Days</dc:creator>
      <category>AI Safety</category>
      <pubDate>Tue, 10 Feb 2026 00:00:00 GMT</pubDate>
      <description><![CDATA[CAI v2 adds explicit anti-deception constraints, support for human oversight in agentic chains, and new evaluation benchmarks. Represents a meaningful update to the corrigibility framing introduced in 2022. <a href="https://zero-days.ai/#news-library">Read all signals →</a>]]></description>
    </item>

    <item>
      <title>KOSA state-level analog passes in 12 states simultaneously — age verification wave accelerates</title>
      <link>https://www.ncsl.org/technology-and-communication/social-media-and-youth-mental-health-legislation</link>
      <guid isPermaLink="true">https://www.ncsl.org/technology-and-communication/social-media-and-youth-mental-health-legislation</guid>
      <dc:creator>Zero Days</dc:creator>
      <category>KOSA / Child Safety</category>
      <pubDate>Thu, 05 Feb 2026 00:00:00 GMT</pubDate>
      <description><![CDATA[Twelve US states enacted minor-protective social media legislation in a single week. Laws vary on verification methods and platform thresholds but share a common requirement: platforms must verify user age before serving algorithmic feeds to minors. <a href="https://zero-days.ai/#news-library">Read all signals →</a>]]></description>
    </item>

    <item>
      <title>FTC enforcement action against AI hiring tool — unlawful discrimination via proxy variables</title>
      <link>https://www.ftc.gov/news-events/news/press-releases</link>
      <guid isPermaLink="true">https://www.ftc.gov/news-events/news/press-releases</guid>
      <dc:creator>Zero Days</dc:creator>
      <category>FTC / US Regulation</category>
      <pubDate>Fri, 30 Jan 2026 00:00:00 GMT</pubDate>
      <description><![CDATA[FTC consent order against HiringAI Inc. for using ZIP code and shopping behavior as proxy variables for race in automated screening tools. First enforcement action under the FTC's updated AI hiring guidance. <a href="https://zero-days.ai/#news-library">Read all signals →</a>]]></description>
    </item>

    <item>
      <title>Academic: AI content moderation at scale — accuracy vs. context tradeoffs across 50M posts</title>
      <link>https://arxiv.org/abs/2603.00000</link>
      <guid isPermaLink="true">https://arxiv.org/abs/2603.00000</guid>
      <dc:creator>Zero Days</dc:creator>
      <category>AI Safety</category>
      <pubDate>Sun, 25 Jan 2026 00:00:00 GMT</pubDate>
      <description><![CDATA[Stanford HAI study finds automated moderation accuracy drops 31% on contextually ambiguous content vs. clear-cut violations. Key finding: context window size and cultural training data are dominant factors — more so than model scale. <a href="https://zero-days.ai/#news-library">Read all signals →</a>]]></description>
    </item>

  </channel>
</rss>