AI regulation and standards

OpenAI describes its approach to text watermarking in response to the EU AI Act, which requires generative AI providers to make generated text identifiable in a machine-readable way. Its textGrain technology adds an invisible statistical signal to word choices, and the detector's error rates, shorter text, and synonym replacement limit reliability, so detector access is initially restricted to approved researchers and expert organizations.

Sen. Adam Schiff, a California Democrat on committees overseeing tech and AI, discusses regulating the tech industry in an interview with The Verge's Decoder. He argues the Trump administration uses national security claims to justify its actions, and contrasts that with a White House event where AI executives signed a nonbinding pledge to evaluate their own AI systems.

Anthropic CEO Dario Amodei proposed a three-step plan for slowing AI development, including embedding third-party evaluators in labs, coordinating across the domestic industry, and forging international agreements, potentially with government assistance. OpenAI CEO Sam Altman, Google DeepMind co-founder Demis Hassabis, and SpaceX CEO Elon Musk publicly appeared to agree with aspects of all three steps.

This article profiles 16 companies building governance, guardrail and trust tools for teams running LLMs in production, listed alphabetically. It covers vendors such as Collibra, Confident Security and Credo AI, with features like trust scores, encrypted inference and policy packs for regulations including the EU AI Act.

House Speaker Mike Johnson cancelled Thursday's votes, sending lawmakers home a day early ahead of the midterm election recess. The change means House members will not vote on Republican Thomas Massie's resolution to impeach Defense Secretary Pete Hegseth.

US Vice President JD Vance has rejected calls for global regulation of AI safety risks. Addressing Dario Amodei, co-founder of Anthropic, who urged Washington to coordinate AI control including with China, Vance said companies building the most advanced models should stop if they are creating a Frankenstein.

More than 20 members of Congress have called for new or stronger AI regulation after Jacob Coxon, a former OpenAI researcher, quit Anthropic and warned that AI could "kill us all by the end of the decade." Several bills are in early stages, including the Frontier Act, the AI Kill Switch Act, and the Ban Artificial Superintelligence Act, and Congress is mostly out of session until the midterms.

Anthropic has explained how it will apply invisible watermarks to Claude-generated text to comply with the EU AI Act's transparency rules. The text marking system is described as a version of SynthID-Text, an open-source watermarking technology from Google DeepMind that creates detectable patterns using wording probabilities.

Anthropic plans to watermark text generated by Claude, starting with a global launch because it has no durable way yet to scope the watermark by region. The watermark is applied during generation by changing the source of randomness in some token choices, rather than adding hidden characters, and it is based on Google DeepMind's SynthID-Text approach. The change is made to comply with the EU AI Act, and models launched before August 2, 2026, fall under the EU's transition period.

The European Union launched a new team in Brussels to monitor AI companies for violations of its AI Act, including sexually explicit material, fake photos and videos, and cyber threats to public infrastructure. The AI Act comes into force on Sunday and requires AI companies to label chatbot output and imagery as AI-generated, and Brussels can fine firms or cut off their access to the EU market for breaches.

AI companies have poured more than $20 million into a New York Democratic primary in Manhattan's 12th congressional district, a race between Alex Bores, Micah Lasher and Jack Schlossberg. Leading the Future, backed by Andreessen Horowitz, Greg Brockman and Perplexity, spent $8 million opposing Bores, while Public First Action, funded with $20 million from Anthropic, has supported him with $11 million. The contest has become a proxy fight over how much government should regulate AI.

The source is a news article describing five AI-specific risk management frameworks that organizations can use to identify AI failure modes, put controls in place, and demonstrate responsible AI use. It notes the frameworks differ in focus, covering governance, technical security controls, threat modeling, or regulatory compliance, and that they overlap in core practices. The excerpt covers ISO/IEC 42001 (published December 2023) and the NIST AI Risk Management Framework (released January 2023).

Big Tech lobbyists have pushed for federal preemption of state AI laws, a single nationwide AI rulebook passed by Congress and signed by the president. Their efforts have faced roadblocks and political backlash, and they risk losing a cooperative Congress if Democrats gain control after the midterms.

OpenAI announced its support for the European Commission's Code of Practice on Transparency of AI-Generated Content, which implements the EU AI Act. The company says its provenance work began in 2024 with C2PA metadata on DALL·E 3 images, and that it now uses C2PA metadata together with SynthID watermarks on OpenAI-generated images, plus a public verification tool at openai.com/verify.

The Verge's Regulator newsletter returns after a two-week break and covers tech politics in Washington, DC. The source text is mostly a subscription pitch, a photo caption and an opening note, and it does not report the specific AI regulation developments the headline refers to.

OpenAI proposed mandatory pre-release evaluations of the most capable AI models by the Center for AI Standards and Innovation (CAISI), while arguing CAISI should recommend mitigations rather than approve or block deployments. The proposal, titled "Democratic Governance of Frontier AI: A blueprint for a federal framework," came a day after a White House executive order on advanced AI innovation and security. It also calls for audits, transparency reports, incident reporting, whistleblower protections, and stronger security controls for frontier AI systems.

Hours before he was set to sign an executive order calling for a government safety review of new AI models before release, Donald Trump abruptly backed out. He said the US would not slow its AI race, citing American dominance and competition with China.

AWS Security Blog announced a new compliance guide, ISO/IEC 42001:2023 on AWS, for organizations building an Artificial Intelligence Management System (AIMS) on AWS services. The guide maps ISO 42001:2023 clauses 4–10 and Annex A controls A.2–A.10 to AWS services and architectural capabilities, and it notes that customers remain responsible for defining their AIMS scope, implementing controls, and demonstrating conformity during certification audits.

CrowdStrike has been selected for OpenAI's Trusted Access for Cyber (TAC) program, which gives verified, selected defenders governed access through identity verification and tiered controls. OpenAI released GPT-5.4-Cyber, a frontier model designed for defensive cybersecurity, alongside the TAC expansion. The article is largely promotional, describing CrowdStrike's Falcon platform and AgentWorks framework and noting that the EU AI Act's next phase takes effect August 2, 2026.

California Governor Gavin Newsom signed an executive order giving the state four months to develop AI policies that prioritize public safety. Companies seeking state contracts must show policies that prevent AI from distributing child sexual abuse material and violent pornography, avoid harmful bias, and prevent unlawful discrimination, detention, and surveillance. The order also directs the state to develop best practices for watermarking AI-generated or manipulated images and videos.