Global Tech Policy Tracker — 2026-03-24

New Legislation & Proposals

United States: Bipartisan AI in Federal Courts Bill

• What it does: Establishes federal oversight and clear guidelines for the use of artificial intelligence systems within US federal courts, addressing a gap in current judicial policy.
• Who it affects: Federal judges, law clerks, legal practitioners, and litigants who interact with AI-assisted tools in the federal court system.
• Timeline: Bill introduced in Congress on March 19, 2026; next steps include committee review and floor votes.
• Key provisions:
  • Mandates scrutiny and disclosure requirements for AI tools used in federal judicial proceedings.
  • Addresses broader concerns about AI reliability and accountability in legal contexts raised on Capitol Hill.
  • Bipartisan in nature, signaling cross-party concern about AI in the justice system.

2 sources

politico.com

politico.com

politico.com

politico.com

European Union: AI Act Omnibus Delay — Fixed 2027/2028 Deadlines

• What it does: European Parliament committees voted 101–9 to postpone AI Act high-risk system compliance deadlines, replacing a flexible European Commission trigger mechanism with fixed statutory dates.
• Who it affects: Companies developing or deploying high-risk AI systems in the EU, including healthcare AI, biometric systems, and critical infrastructure tools; also general-purpose AI model providers.
• Timeline: The committee vote has now passed; the report will proceed to a full Parliament plenary vote. New fixed deadlines are set for December 2027 and August 2028.
• Key provisions:
  • High-risk AI system obligations pushed to December 2027 (first tranche) and August 2028 (second tranche).
  • Replaces the Commission's discretionary trigger with hard legislative dates, giving industry firmer planning horizons.
  • Part of the broader AI Act Omnibus simplification package, which also aims to reduce compliance burdens on SMEs and general-purpose AI model providers.

1 sources

ppc.land

ppc.land

Regulatory Actions

Civil Society vs. EU Commission: Criticism Over AI Act Dilution

• Action: Policy analysts and civil society groups published a critical assessment on March 18, 2026, warning that the European Commission's omnibus proposal to simplify the AI Act risks "watering down" fundamental rights protections that were core to the original regulation. As the omnibus proposal moves toward trilogue negotiations between Parliament, Council, and Commission, critics argue that deregulatory impulses are overriding the Act's original goals.
• Significance: The critique highlights a growing tension in Brussels between the EU's competitiveness agenda (streamlining compliance for industry) and its stated commitment to protecting citizens from AI harms. The outcome of trilogue negotiations will set the tone for AI governance across the bloc — and potentially influence how other jurisdictions model their own frameworks.

Industry Response

The formal passage of the EU Parliament committee vote on fixed AI Act deadlines is being closely watched by compliance teams. Legal analysis published this week by Prokopiev Law highlights that the EU Commission's original Omnibus I proposal (adopted November 2025) had already aimed to reduce compliance burdens on AI developers, deployers, and providers of general-purpose AI models — and the Parliament's decision to set fixed dates (rather than leaving the trigger to Commission discretion) is now seen as providing more legal certainty for enterprises planning multi-year compliance roadmaps.

Separately, SME-focused commentary published this week notes that many medium-sized businesses remain dangerously underprepared for the EU AI Act's August 2026 obligations — particularly those using AI embedded in standard commercial software, which can unexpectedly trigger high-risk classification and potential fines of up to €35 million or 7% of global annual turnover.

What This Means

• The US federal AI-in-courts bill signals that Congress is moving beyond broad AI strategy documents and into sector-specific legislation. AI developers building tools for legal research, case analysis, or courtroom assistance should watch this bill closely — it may set disclosure and accountability precedents that ripple into other high-stakes domains.

• The EU Parliament's 101–9 committee vote in favor of fixed AI Act deadlines gives industry far more planning certainty than the original flexible trigger, but it also locks in a two-year extension window during which enforcement will be limited. Startups and scale-ups operating in EU markets should use this window to build compliance infrastructure rather than treat it as a reprieve.

• The omnibus simplification debate is not over. Trilogue negotiations between Parliament, Council, and Commission remain ahead, and the gap between civil society's demands for stronger protections and industry's push for lighter rules will be a defining political contest in Brussels for the remainder of 2026.

• SMEs face an often-overlooked trap: AI embedded in standard off-the-shelf software can qualify as high-risk under the EU AI Act. The August 2026 obligations deadline has not moved, and businesses that assume the deadline extension covers them may face significant regulatory exposure.

Region Scorecard

Coverage period: March 17–24, 2026. All claims are sourced from verified publications within this window. Sections with insufficient fresh data have been omitted.