Colorado’s SB 24-205, the first comprehensive state AI law in the United States, reaches its scheduled effective date today already repealed, court-stayed, and quietly disowned by the attorney general who once championed it. From passage to collapse: 14 months.

The sequence is worth tracing because it now functions as a template. On April 9, Elon Musk’s xAI sued Attorney General Phil Weiser, arguing the statute’s anti-disparate-impact provisions compelled speech and embedded “state-enforced orthodoxy” on DEI. Fifteen days later the Justice Department intervened, citing Executive Order 14365 in what the agency described as its first-ever intervention against a state AI law. Harmeet K. Dhillon, DOJ Civil Rights chief, said the department “will not stand on the sidelines while states such as Colorado coerce our nation’s technological innovators.”

Weiser didn’t fight. On April 27, Magistrate Judge Cyrus Y. Chung approved a joint xAI–DOJ–Weiser motion suspending enforcement, citing pending legislative revisions. The attorney general had effectively joined the plaintiffs against his own statute. The AG’s office now references SB 24-205 only in the past tense and points the public to its replacement.

That replacement, SB 26-189, was signed by Governor Jared Polis on May 14 and takes effect January 1, 2027. It’s a narrower ADMT framework, minus the bias audits and mitigation plans that made the original ambitious in the first place. xAI retains a 28-day window to refile for injunction after final AG rulemaking, a leash on the new law before it even arrives.

The legible read for other statehouses: a coordinated industry-plus-DOJ challenge, paired with a governor willing to repeal-and-replace, can dismantle a state AI regime before it regulates a single algorithm. Colorado was the test case. It didn’t survive its own start date.

Sources

Sources