Urges generative artificial intelligence companies to make voluntary commitments regarding employee whistleblower protections.

This New Jersey Assembly Resolution (AR 158) urges generative AI companies like OpenAI, Anthropic, and Google to voluntarily commit to protecting employees who report safety concerns. It asks companies to: establish anonymous channels for employees to raise risk-related issues with boards or regulators, prohibit retaliation for speaking out, and provide legal and technical safe harbor for employees conducting good-faith safety evaluations without fear of account suspension or legal action. The resolution directly affects current and former employees of AI companies who might want to disclose risks like misinformation, bias, or system flaws. It is not a binding law but a call for companies to adopt specific protections to enable safer AI development, emphasizing that employees possess critical knowledge about AI risks but face barriers like confidentiality agreements.