Welcome to This Week in AI Regulations for March 29, 2026.

In the United States, the Federal Reserve has updated its supervisory approach to better enable responsible innovation in digital assets and artificial intelligence. The Federal Reserve rescinded certain crypto-related supervisory letters and replaced restrictive policy statements with ones facilitating innovation. Key requirements include that banks proactively implement governance, risk management, and oversight policies for AI usage. Banks must also understand risks and legal obligations in bank-fintech partnerships.

Also in the United States, the Securities and Exchange Commission reported an investigation revealing unauthorized audio recording and use of an unapproved AI transcription service by a former SEC attorney, resulting in mishandling of nonpublic information. The SEC emphasized the need for strict controls on information handling, requiring consent before recording interviews, use of only approved applications for recording and transcription, and protection of nonpublic information from unauthorized transmission.

Moving to Australia, new national expectations have been introduced for data centre and AI infrastructure developers. These five expectations focus on prioritizing Australia’s national interest, supporting the energy transition, and ensuring water sustainability. The policy impact is to adopt and align with these national expectations for data centre and AI infrastructure projects.

In the European Union, several regulatory updates are underway. The European Parliament is examining proposals to amend existing EU regulations on AI implementation alongside food and feed safety simplification and strengthening.

Additionally, the Law Enforcement Working Party meeting held on March 26, 2026, included presentations and exchanges on key law enforcement topics and draft European Commission guidelines on high-risk AI systems. Engagement with initiatives on human trafficking, migrant smuggling, and trafficking of firearms and explosives was also part of the agenda.

Further, the European Union introduced specific adjustments to simplify compliance and reduce administrative burdens while maintaining fundamental rights protection under the General Data Protection Regulation and the AI Act.

In Germany, a trilateral pilot project simulating an AI real laboratory has been successfully completed. This project produced a detailed roadmap integrating the EU AI Regulation and Medical Device Regulation requirements. The German Federal Network Agency, Bundesnetzagentur, is mandated to establish a national AI real laboratory by August 2, 2026.

In Israel, a regulatory pilot program has been launched offering financial support and regulatory guidance for AI healthcare technologies with high autonomy levels. This sandbox program focuses on challenges related to automation and delegation of clinical tasks. Applicants must be industrial corporations with research and development projects at Technology Readiness Level 6 or above. Projects must address at least one specified challenge, and applications must be submitted with complete documentation via the online personal area.

That wraps up today's regulatory updates. Visit carveragents.ai for more information.