Nebraska bill would force major AI developers to publish safety plans and report incidents

Sen. Tanya Storer introduced LB 10-83, the Transparency in Artificial Intelligence Risk Management Act, saying the measure would require the largest AI model developers and chatbot providers to publish safety plans, disclose risk assessments and report safety incidents to the attorney general.

"We are focused on consumer-facing conversational AI systems that minors are likely to access," Storer said, describing an amendment (AM 2025) meant to exclude routine customer-service bots and internal tools. She said the bill applies to companies above high revenue thresholds and is designed as a light-touch transparency framework rather than a technical mandate.

Witnesses including Andrew Doris of the Secure AI Project and several child-safety advocates told the committee the bill balances the need for regulatory visibility with flexibility for developers: companies write their own safety plans so regulators and the public can hold them to those commitments rather than prescribing specific technical solutions. Doris said the approach helps reconcile fast-moving industry practices with public oversight.

Testimony also covered operational and constitutional questions: committee members asked whether delegating definitional authority to the attorney general constitutes an unlawful delegation, and several asked for clarity on the proposed fund and its intended uses. Supporters said the bill includes guardrails and that an updated fiscal note was expected. No committee vote was taken at the end of the hearing.