A Resolution designating August 31, 2025, as "Overdose Awareness Day" in Pennsylvania.

2025-2026 Regular Session Introduced by Rosemary Brown and 12 co-sponsors

Urges major generative AI firms to voluntarily adopt employee protections and whistleblower safeguards, enabling anonymous risk reporting and public evaluation of safety concerns.

Referred to Rules & Executive Nominations

Summary — SR 121

Title: Urges generative artificial intelligence companies to make voluntary commitments regarding employee whistleblower protections
Classification: Senate Resolution (non‑binding)
Introduced: February 18, 2025
Status: Introduced; referred to Senate Labor Committee
Primary Sponsors: Kenya Wicks; Brian Strickland; Bill Cowsert; Bo Hatchett; Harold Jones II; Kay Kirkpatrick; San Buenaventura; Beth Mizell
Cosponsor: Gerald Boudreaux
Related: SCR 151 (companion), AR 158 (companion)

Purpose / Intent

SR 121 asks major generative artificial intelligence (AI) companies to voluntarily adopt a set of employee protection and transparency commitments intended to encourage internal and independent reporting of risk‑related concerns about AI systems. The resolution aims to reduce retaliation risk for employees and independent evaluators so that safety issues, biases, misinformation risks, and other harms are more readily identified and mitigated.

Key provisions (voluntary principles the resolution urges companies to adopt)

The resolution asks generative AI companies to commit to the following principles:

Not entering into or enforcing agreements that bar disparagement or criticism about risk‑related concerns, and not retaliating by withholding vested economic benefits for such criticism.

Providing a verifiably anonymous process for current and former employees to raise risk‑related concerns to the company board, regulators, or an appropriate independent organization.

Supporting a culture that permits employees and former employees to raise risk‑related concerns publicly (to the extent trade secrets and IP are protected).

Offering legal and technical “safe harbor” for good‑faith system evaluation — protecting researchers from legal reprisal, account suspension, or termination while still protecting trade secrets and IP.

Not retaliating against employees who publicly disclose risk‑related confidential information after internal processes have failed.

Allowing employees to continue to report publicly until adequate anonymous reporting channels exist, with guidance to avoid unnecessary disclosure of confidential information.

The resolution also directs that copies be transmitted to CEOs of leading generative AI firms (examples listed: OpenAI, Anthropic, Google, Inflection, Meta, Midjourney, Cohere).

Who is affected

Primary targets: Generative AI companies and their leadership (corporate governance).

Direct beneficiaries: Current and former employees, internal safety teams, independent evaluators/researchers.

Secondary: Regulators, independent safety organizations, and the general public (through potential reductions in AI risk).

Procedural / legal character & likely impact

SR 121 is a non‑binding, symbolic resolution that urges voluntary corporate action rather than imposing legal obligations.

If companies adopt these commitments, potential impacts include increased internal reporting, more independent safety testing, and a stronger whistleblower culture.

Limitations: Because the resolution does not change statutory law, concepts like “legal safe harbor” would generally require future legislation or contract changes to create enforceable protections; trade‑secret and IP protections remain a key balancing issue.

The resolution may serve as a catalyst for corporate policy changes or follow‑on legislation at state or federal levels.