Researchers Fault OpenAI on Illinois AI Audit Law
Researchers claim OpenAI's policy head mischaracterized Illinois AI audit rules as basic safety checks.
Miles Brundage, Nathan Calvin, and Steven Adler posted criticism of OpenAI's Head of State Policy comments on an Illinois AI audit law. They said the law calls for procedural compliance audits rather than checks on whether models are safe. Calvin compared the OpenAI description to equating the requirements with vehicle brake checks. Brundage added that the comments portrayed the rules as inconsistent with existing frameworks, though they follow a standard copy-and-paste approach with some additions.
Combined views
14.3K
7 posts, first seen 3d ago
Reactions from ranked influencers
7 posts
OpenAI not have a misaligned policy team, challenge level impossible
These comments from OpenAI in Bloomberg about existing and proposed AI state auditing requirements don't make sense. Its not true that the Illinois audit requirements are the equivalent of "making sure the brake lights and windshield wipers are working." The Illinois audit…
Disappointing since OAI has been way better on this front lately overall, but the rot runs deep + employees on other teams sadly need to watch this stuff like a hawk
Besides what Nathan said, I'll add that the OAI comments give the sense that this is some random, totally inconsistent framework from what already exists. No - as is the standard these days, it is a "copy/paste plus some new stuff" job, precisely to avoid inconsistency.
I’m very confused: It seems like OpenAI’s Head of State Policy mischaracterized the Illinois AI audit law. Was he mis-summarized? According to METR, the law doesn’t audit whether models are safe (like a brake check), it is just a procedural compliance audit.
OpenAI Global Affairs put out a new statement today making clear that they don't think current state laws like SB 53 in CA and SB 315 in IL are sufficient and that its important for legislatures to revisit legislation in light of lessons from real world events like the HF…
These comments from OpenAI in Bloomberg about existing and proposed AI state auditing requirements don't make sense. Its not true that the Illinois audit requirements are the equivalent of "making sure the brake lights and windshield wipers are working." The Illinois audit…
Glad to see this clarification!
OpenAI Global Affairs put out a new statement today making clear that they don't think current state laws like SB 53 in CA and SB 315 in IL are sufficient and that its important for legislatures to revisit legislation in light of lessons from real world events like the HF…