After federal deregulation took hold in January, states are moving aggressively to fill the regulatory void. This week saw multiple developments and a clear signal from the Trump administration that it expects industry to lead, not government.
1. Colorado AI Act Implementation Approaches
The news: Colorado’s Consumer Protections for Artificial Intelligence Act (SB 24-205), signed into law in May 2024, is the first comprehensive state AI bias law. It requires developers of “high-risk AI systems” (used in hiring, lending, benefits, housing) to:
- Conduct impact assessments and bias testing
- Provide transparency disclosures to consumers
- Report to the Colorado Attorney General
- Implement risk management policies
Why it matters: The law was originally set to take effect February 1, 2026, but the implementation deadline was pushed to June 30, 2026 by the AI Sunshine Act (SB 25B-004). Companies are now preparing for compliance. If enforced as written, Colorado sets a precedent that other states may follow.
Status: Implementation delayed to June 30, 2026.
2. Massachusetts Debates AI Regulation
The news: The Massachusetts legislature is advancing multiple AI-related bills, including the AI Disclosure Act (H.81) and proposals regulating AI use in healthcare and consumer targeting. Some proposals echo elements of California’s vetoed SB 1047, particularly around transparency and safety requirements.
Why it matters: Massachusetts is a major tech hub (Boston, Cambridge). Strict rules here could influence hiring and model development decisions industry-wide.
Key debate: State officials vs. tech companies. State says “safety first”; tech says “innovation will move to other states.”
3. Trump Administration Signals State Preemption
The news: Building on Executive Order 14179 (Removing Barriers to American Leadership in AI, signed January 23, 2025), the Trump administration issued an executive order in December 2025 on “Ensuring a National Policy Framework for Artificial Intelligence.” It created a DOJ AI Litigation Task Force specifically to challenge state AI laws in federal court.
Why it matters: This is a direct challenge to states like Colorado, California, and New York that are pursuing their own AI regulation. It signals potential legal battles over whether the Commerce Clause can preempt state AI regulation.
Legal question: Can the federal government preempt state AI regulation? This remains unresolved and is likely heading to the courts.
4. EU AI Act Enforcement Update
The news: The EU AI Act’s prohibited-practices provisions (Article 5) took effect on February 2, 2025, banning AI systems used for social scoring, untargeted facial recognition scraping, and emotion inference in workplaces and schools. The next major milestone — obligations for high-risk AI systems under Annex III — is set for August 2, 2026.
Why it matters: This is the world’s most comprehensive AI regulation. Companies serving EU customers must comply or face penalties up to 7% of global annual turnover (EUR 35 million) for prohibited-practice violations. Even U.S. firms are preparing, as the Act creates de facto global standards.
Impact on U.S. companies: Firms serving EU customers must comply. A European Digital SME Alliance survey found over 60% of small and medium-sized tech companies say they are not adequately prepared.
5. China Accelerates Domestic AI Development
The news: Chinese AI labs released a wave of major models around the Lunar New Year period in February 2026, including Alibaba’s Qwen 3.5 (native multimodal capabilities), ByteDance’s Doubao 2.0 (complex reasoning), Zhipu AI’s GLM-5 (agentic intelligence), and Moonshot AI’s Kimi K2.5. China’s 15th Five-Year Plan (2026–2030) highlights AI as a national priority.
Why it matters: As the U.S. shifts to deregulation and the EU imposes strict rules, China is building AI at scale. China does have its own AI regulations (including the 2023 Interim Measures for Generative AI Services), but they are oriented more toward content control than safety testing.
Geopolitical context: The U.S. faces a three-way dynamic: compete with China, navigate the EU’s strict rules, and manage internal fragmentation.
Analysis: The State vs. Federal Stalemate
The clear pattern this week: the U.S. is entering a regulatory stalemate.
- States are moving because the federal government has stepped back (Trump’s deregulation)
- The federal government wants states to step back because uniform regulation is easier for business
- Tech companies want federal preemption, but they’re getting neither federal rules nor permission to ignore states
The likely outcome: A multi-year fight. Either:
- The federal government establishes preemption authority and writes national AI rules (unlikely under Trump), or
- States win and companies adapt to a patchwork of rules (expensive but survivable), or
- A compromise emerges where federal law sets a floor and states can exceed it (like environmental regulation)
For developers: Build with compliance in mind. Assume Colorado, California, Massachusetts, and New York rules will apply to your system if you want to operate nationwide.
What to Watch Next Week
- EU enforcement updates — Ongoing implementation of prohibited-practices rules
- Colorado and Massachusetts — Movement on state AI legislation
- Trump administration guidance on AI export controls and semiconductor restrictions
- Congressional hearings on AI liability and insurance (potential federal action)