Comment by Rumman Chowdhury

Critically, robust regulation of high-risk AI does not stifle innovation; it enables it. In safety engineering there is a saying: brakes help you drive faster. The existence of reliable brakes does not slow cars down—brakes create the conditions under which drivers can safely go farther, faster, with confidence. The same is true here: clear guardrails, mandatory audits, and user rights give both companies and the public confidence that AI can be used in certain domains without becoming an uncontrolled experiment on people’s lives. The bill’s requirement for independent, periodic third-party audits for high-risk AI systems is particularly important. The most meaningful improvements have come when evaluations are run by entities structurally independent of product and revenue teams and when results are tied to concrete risk-mitigation obligations. Provisions such as independence rules for auditors, regular audits for both developers and deployers, and a public database of redacted reports move New York toward a model of credible evaluation and public accountability.