US vs EU on AI rules
The White House has released a national AI policy framework and is pressing Congress to pass a federal law that would pre‑empt state rules, focusing on child safety, consumer protection and workforce training. (pymnts.com) At the same time EU Parliament committees voted to delay compliance deadlines for ‘high‑risk’ AI systems with fixed enforcement dates in December 2027 and August 2028, while keeping heavy liability and fines—up to €35 million or 7% of global turnover—raising cross‑border compliance headaches for multinationals. (ppc.land), (xpert.digital)
The White House published its legislative framework on March 20, 2026, organized around six policy objectives that the administration says should guide congressional AI legislation. (whitehouse.gov) The framework explicitly seeks streamlined permitting so data centers can generate power on site and aims to prevent ratepayers from absorbing data‑centre energy costs. (whitehouse.gov) It proposes a dual approach to intellectual property and model training that pairs stronger protections for creators with allowances for “fair use” to enable AI improvement, and it includes explicit language on safeguarding free‑speech protections from government censorship via AI. (whitehouse.gov) Earlier congressional maneuvering included a mid‑2025 effort to insert a multi‑year moratorium on state AI laws into budget reconciliation language, a provision that the Senate later stripped from the measure. (en.wikipedia.org) State attorneys general and a bipartisan group of officials previously pressed Congress to reject a long moratorium on state regulation, while legal analysts warn that any federal effort to block state laws or to withhold federal funds could prompt federal‑court challenges. (pymnts.com) In Brussels, the European Parliament’s IMCO and LIBE committees voted on March 18, 2026 by 101‑9 to set fixed application dates for high‑risk AI obligations — 2 December 2027 for Annex III systems and 2 August 2028 for Annex I safety‑component systems. (ppc.land) The committees kept the AI Act’s enforcement architecture including administrative fines of up to €35 million or 7% of global annual turnover for the most serious breaches under Article 99. (ai-act-service-desk.ec.europa.eu) Parliamentary and Council sign‑off remains required before the changes become law, and industry groups welcomed the fixed delays as clarity for compliance timelines while consumer and rights organizations warned the pause could weaken protections. (europarl.europa.eu)
