AI Regulation & Policy

Today's Stories

1. 1

   DeepMind CEO urges U.S.-led oversight of advanced AI

   DeepMind CEO Demis Hassabis has called for U.S.-led oversight of advanced AI models, according to reporting by Yahoo Finance Tech Editor Dan Howley. The call for government oversight from a leader at one of the world's most prominent AI research organizations signals that industry figures believe regulatory frameworks are needed as AI systems become more powerful. This reflects ongoing debate over how advanced AI should be governed.

   How U.S. policymakers respond to this call and whether other AI leaders join Hassabis in advocating for centralized oversight mechanisms.

2. 2

   NYC and EU tighten AI hiring rules; 95% of employers lack required audits

   NYC's Local Law 144 enforcement has been active since July 2023, with the Department of Consumer and Worker Protections (DCWP) issuing first penalties in Q4 2025 and shifting to proactive investigations in January 2026. The EU AI Act's Article 50 transparency obligations take effect August 2, 2026, and Annex III high-risk AI obligations take effect December 2, 2027. A NYS Comptroller audit found 17 potential violations that DCWP missed. Only ~5% of NYC employers using AI hiring tools publicly post required bias audit results; the other 95% are now exposed to proactively-enforcing regulators. Penalties for violations start at $500 and rise to $1,500 per day for subsequent violations across multiple violation types. Qualified independent auditors are booked months in advance, and reconstructing missing historical hiring data takes 60–90 days—delays that leave employers vulnerable as enforcement intensifies.

   EU AI Act Article 50 transparency obligations are due August 2, 2026 (14 days remaining as of the article's publication). Annex III high-risk AI obligations are due December 2, 2027. Employers must complete an independent bias audit within the previous 12 months, post a bias audit summary on their website, and provide candidate notice at least 10 business days before using an AI employment decision tool (AEDT).

3. 3

   Palantir, Nvidia team up on sovereign AI for governments

   Palantir Technologies announced a partnership with Nvidia to deploy sovereign AI solutions for government agencies, using Nvidia's open AI model Nemotron as a core building block within Palantir's software platforms. The collaboration targets governments' need to keep sensitive data under local control while meeting compliance and national sovereignty requirements—a central requirement for public sector AI adoption that sits at the heart of Palantir's government business.

   The partnership may influence how Palantir approaches new government bids and technology alignment with public sector clients, and could shape how investors view Palantir's role in the wider AI supply chain spanning chips, models, and application-level software.

4. 4

   China launches parallel AI order with 29 nations, offers 5,000 training slots

   Xi Jinping announced 5,000 AI training slots for Global South countries over the next five years at the World AI Conference in Shanghai. A day earlier, 29 nations formally established the World Artificial Intelligence Cooperation Organization (WIKO), headquartered in Shanghai, with founding members including Russia, Brazil, South Africa, Pakistan, and Indonesia—but no Western countries. This is China's clearest move to build an AI governance structure independent of Western influence, anchored in developing-world alliances. Xi also called for AI to remain under human control and pushed back against broad national security justifications in AI policy, a criticism aimed at US export controls on AI chips and technology.

   China's "Smart Economy," spanning AI and other digital technologies, is now worth over one trillion renminbi, roughly $140 billion（約22兆円）, according to Xi—a measure of the scale of China's AI infrastructure and ambitions within this new framework.

5. 5

   McLaren Construction deploys FieldAI robots on UK sites for quality checks

   McLaren Construction has partnered with FieldAI, a physical AI developer, to deploy autonomous quadruped robots across its UK construction sites. The robots will initially capture 360° imagery, generate point cloud data, conduct safety patrols, and verify progress against design models. The robots perform AI-enabled deviation analysis to compare site conditions against design specifications, catching quality issues earlier and reducing rework—a significant shift from remote-controlled or pre-programmed machines. This marks FieldAI's entry into the UK market and extends its existing deployments across hundreds of sites in Europe, Asia, and North America.

   McLaren expects the partnership to deliver more reliable project monitoring and stronger evidence for compliance and quality assurance. Over time, the robots' capabilities will expand to site logistics, dexterous manipulation, and multi-robot coordination as FieldAI's general-purpose systems grow more capable.

6. 6

   Brex builds AI agent control layer at network level, not in rules

   Brex created CrabTrap, an open-source HTTP/HTTPS proxy that intercepts all network traffic from AI agents, examines policy rules, and uses an LLM-as-a-judge to approve or deny requests. The company found that traditional guardrails could not contain what agents were doing with real credentials like API keys and OAuth tokens. Brex's approach addresses a gap in how AI agents are currently governed—frameworks like OpenClaw enable agents to act, but lack enterprise-scale safeguards. By enforcing policy at the network layer rather than in the agent's code, organizations can audit and control agent behavior in real time, even when agents have genuine credentials to systems that matter.

   Brex CEO Pedro Franceschi frames this as a shift in how IT leaders should think about agent governance: moving from SDK-level permissions and model guardrails to centralized network control. How widely CrabTrap is adopted, and whether other enterprises adopt similar network-layer enforcement, will signal whether this architectural approach becomes standard practice.