AI developers pledge enhanced safety controls amid US legislative push

AI regulation internal controls recursive self-improvement AI safety legislation liability

Major AI developers, including Google, Anthropic, OpenAI, and xAI, have committed to strengthening internal controls for their models following a meeting with President Trump. The agreement focuses on tracking model capabilities during training and deployment, with specific emphasis on preventing risks in biosafety, cybersecurity, and chemical threats. To ensure accountability, these companies have pledged to employ independent external auditors to evaluate their safety systems.

In the legislative sphere, Representative Ro Khanna has introduced the 'Human Control Over AI Act.' This bill seeks to prohibit recursive self-improving AI models—systems that can autonomously enhance their own architecture—until federal safety standards and a dedicated regulatory agency are established. The proposed law includes severe penalties, including criminal charges for deploying AI that causes mass civilian harm or for employees who bypass safety mechanisms.

Concurrently, Vice President JD Vance has cautioned against granting AI firms liability protections similar to Section 230, arguing that developers must remain responsible for their products. Vance also warned against 'cartel-like' behavior among leading AI companies, emphasizing the administration's desire to maintain market competition while balancing innovation and safety.

Entities

Anthropic · Bernie Sanders · Donald Trump · Greg Casar · JD Vance · OpenAI · Ro Khanna · U.S. House of Representatives

Claims

What the coverage asserts, and how many sources carry each claim.

[○ 1 SOURCE] AI developers committed to using independent external auditors to assess the effectiveness of their safety controls. sputniknews.cn
[○ 1 SOURCE] Under the EU AI Act, the duty regarding AI literacy became effective on February 2, 2025. digitalworkplacegroup.com
[○ 1 SOURCE] Internal control mechanisms will address biosafety, cybersecurity, and the prevention of chemical threats. sputniknews.cn
[○ 1 SOURCE] JD Vance argued that AI companies should not receive liability shields like Section 230 protections. www.mangalorean.com
[○ 1 SOURCE] Khanna's proposed law includes criminal penalties for deploying AI that causes mass civilian harm or for employees who bypass safety protocols. www.ithome.com
[○ 1 SOURCE] The US administration seeks to prevent major AI developers from forming cartels that weaken competition. www.mangalorean.com
[○ 1 SOURCE] AI companies including Google, Anthropic, OpenAI, and xAI signed an agreement to implement internal controls during model training and deployment. sputniknews.cn
[● 2 SOURCES] Rep. Ro Khanna introduced legislation to ban recursive self-improving AI models until federal safety safeguards and a regulatory agency are established. cryptobriefing.com · www.ithome.com