Top Stories – May 8, 2026
EU Council and Parliament Agree to Simplify and Streamline EU AI Act Rules Tier 2
The EU Council and Parliament reached a provisional agreement on May 7 to amend the EU AI Act, extending the AI regulatory sandbox deadline to August 2027 and clarifying which body – the EU AI Office or national authorities – has jurisdiction over different AI system types.
Trump’s CAISI Signs Pre-Release AI Testing Agreements with Google, Microsoft, and xAI Tier 2
The Center for AI Standards and Innovation announced voluntary agreements allowing U.S. government evaluators to test AI models from Google DeepMind, Microsoft, and xAI before public release, marking the first concrete step toward federal AI model oversight under the Trump administration.
White House Weighs Mandatory Pre-Release Vetting of New AI Models Tier 3
The Trump administration is considering an executive order that would require government vetting of new AI models before release, driven by national security concerns over Anthropic’s Mythos cybersecurity model – a notable reversal from its earlier deregulatory position on AI oversight.
Scoring: Tier 1 = regulatory changes/enforcement/EOs | Tier 2 = state/international/corporate | Tier 3 = analysis/commentary. Impact criteria: regulatory intent, scale, timeline.