NJ SR121

New Jersey Senate resolution calling on generative AI companies (including OpenAI, Anthropic, Google, Meta, Midjourney, Inflection, and Cohere) to voluntarily adopt employee whistleblower protections

Companies would commit not to enforce non-disparagement agreements or retaliate against employees who raise risk-related concerns about AI technology

Companies would establish anonymous reporting processes for current and former employees to report safety concerns to company boards, regulators, and independent organizations with relevant expertise

Companies would provide both legal and technical safe harbor for independent researchers conducting good-faith system evaluations, protecting them from lawsuits, account suspension, or termination

Employees would be permitted to publicly disclose risk-related confidential information after internal processes have failed, while still protecting trade secrets and intellectual property