The core development is the rapid evolution of enterprise AI contracts, which incorporate AI-specific clauses for data usage, testing, transparency, incident response, audit rights, and liability allocation—outpacing slow legislation like Colorado's AI Act (effective June 30, 2026), Illinois's employment AI law (January 1, 2026), and others in Texas and California imposing risk management on high-risk systems.[2][3][4][5] A December 2025 Executive Order (EO) under the Trump Administration directs federal agencies, including a new AI litigation task force led by the Attorney General, to challenge "onerous" state laws via lawsuits, funding conditions, and preemption evaluations due by March 11, 2026; it carves out child safety and infrastructure but signals disruption without immediate federal uniformity.[1][4][5] Involved parties include states (Colorado, Illinois, Texas, California, Utah), federal entities (DOJ, FTC, Commerce, FCC), and companies negotiating vendor contracts for agentic AI risks like autonomous errors.[3][5]
This stems from 2025's state AI statutes creating compliance burdens for startups and enterprises, contrasted with EU AI Act phases (GPAI obligations since August 2025, high-risk full effect August 2, 2026), driving contractual workarounds; no courts have ruled on agentic AI liability yet.[1][3][5][6] Timeline: State laws activate early 2026; EO challenges imminent; contracts adapt now.[2][4]
Newsworthy as legal risks shift to contracts over statutes in 2026's "patchwork" environment, urging in-house counsel to lead governance, documentation, and indemnification before federal-state clashes resolve—especially with Gartner forecasting 80% of firms mandating AI policies.[2][5][8]