AI Regulation & Policy

Today's Stories

1. 1

   Trump administration clamps down on Anthropic's AI model 'Mythos,' with Amazon's CEO reportedly playing a key role

   The Trump White House moved to restrict Anthropic's 'Mythos' AI model — one of the most capable AI systems built to date — in what insiders describe as a pivotal moment for AI regulation in the U.S. According to Fortune, the intervention was partly set off by a phone call from Amazon CEO Andy Jassy, and the fallout is reshaping the rules governing what powerful AI companies can and cannot do. Anthropic, which is partly owned by Amazon and valued at hundreds of billions of dollars, now finds itself at the center of a broader government debate about national security and AI oversight.

   This sets a precedent: if the government can step in to restrict or control one major AI company's flagship model, it could do the same to others — affecting which AI tools businesses and consumers can actually access.

2. 2

   Europe's biggest retailers ask the EU to exempt AI-made ads from its new labeling law

   Eurocommerce, a trade group whose members include Amazon, H&M, IKEA, and Inditex (owner of Zara), sent a letter to the EU's top tech regulator asking that AI-generated advertisements be excluded from the EU AI Act's disclosure requirements. The EU AI Act — set to take effect on August 2, 2026 — requires companies to clearly label content created or altered by AI if it qualifies as a 'deep fake' (a convincing fake image, video, or audio). The retailers argue that AI-made ads designed to sell products, not to deceive, should not fall under that definition.

   If the exemption is granted, shoppers across Europe could continue seeing AI-generated product images and videos in ads without being told they are AI-made — making it harder to know when what you're looking at is real.

3. 3

   Corporate boards are dangerously unprepared for AI, KPMG's global risk chief warns

   KPMG's global head of risk published a survival guide this week warning that most corporate boards — the groups of senior leaders who oversee large companies — were built for a world of slow, predictable decisions, not the fast-moving, hard-to-predict nature of AI. AI systems can act autonomously, make probabilistic (probability-based, not certain) decisions, and are already embedded in core business operations at many companies, yet most boards lack the tools to properly oversee them. KPMG and business school INSEAD have released a set of governance principles aimed at helping boards catch up.

   If the company you work for or invest in has poor AI oversight at the board level, it faces higher risks of costly AI failures, legal liability, or reputational damage — all of which can affect employees, customers, and shareholders.

4. 4

   OpenAI hires a Transformer co-inventor and a Trump-era AI policy official ahead of its stock market debut

   OpenAI, the company behind ChatGPT, hired Noam Shazeer — one of the original inventors of the Transformer (the underlying architecture that powers nearly all modern AI systems, including ChatGPT) — away from Google DeepMind. In the same week, it also brought on Dean Ball, a former AI policy official from the Trump administration. These hires signal that OpenAI is simultaneously racing to strengthen its technical capabilities and build political relationships as it prepares for an IPO (initial public offering, meaning the company will sell shares to the public on a stock exchange).

   OpenAI adding top technical and political talent before going public suggests the company is positioning itself to stay ahead of both its AI rivals and future government regulation — which will shape how ChatGPT and similar tools are allowed to operate.

5. 5

   ServiceNow quietly positions itself as the go-to platform for managing AI across large companies

   ServiceNow, a software company best known for automating IT helpdesks and business workflows, is increasingly being used by large enterprises to oversee and coordinate their AI tools — a role analysts are calling 'AI governance orchestration' (think of it as a control tower that tracks and manages all AI activity across a company). As businesses deploy more AI agents (software that can take actions on its own, like booking meetings or processing invoices), they need a single system to monitor, audit, and control them. ServiceNow is quietly filling that gap.

   Workers at large companies that use ServiceNow may find their employers are using it to track and audit how AI tools are used in their day-to-day work — including what decisions AI makes on their behalf.