Public Policy & Governance
Top Line
The Pentagon has formally designated Anthropic a supply-chain risk after failed negotiations over AI usage guardrails, marking the first time a US company has received such a designation typically reserved for foreign adversaries, with Anthropic vowing to challenge the decision in court.
The US House Energy & Commerce Committee advanced a package of child online safety bills, despite warnings from civil society groups that privacy protections are constitutionally questionable and may not achieve their stated aims.
The UK government is delaying decisions on AI copyright rules following backlash from creative industries, while a House of Lords committee warns against sacrificing the UK's creative sector for speculative AI gains.
The US is reportedly drafting sweeping new chip export controls that would require countries to invest in America in exchange for advanced semiconductors, potentially requiring permits for all Nvidia and AMD global AI chip sales regardless of destination country.
OpenAI lifted its ban on military use and released new financial-services tools and GPT-5.4 with native computer control capabilities, directly filling the gap created by Anthropic's Pentagon standoff while facing significant user backlash.
Key Developments
Pentagon Formally Designates Anthropic Supply-Chain Risk, Company Vows Legal Challenge
The Department of Defense has formally notified Anthropic that it considers the AI company and its products a supply-chain risk, escalating a weeks-long dispute over acceptable use policies for military applications. According to The Wall Street Journal and Bloomberg, this marks the first time a US company has received such a designation, which has historically been reserved for foreign firms with ties to US adversaries like China's Huawei. Anthropic CEO Dario Amodei stated the company has "no choice but to fight this in court," while claiming the designation will have limited impact on the "vast majority" of customers outside the defense sector. Talks between Anthropic and the Pentagon reportedly resumed after initial breakdown, though the formal designation proceeded regardless.
The dispute centers on Anthropic's refusal to remove restrictions against using its Claude AI for surveillance and autonomous weapons systems. The Pentagon's move comes as OpenAI simultaneously announced it is lifting its ban on military use and releasing new tools specifically for defense applications, directly filling the market gap. The contrast has prompted criticism from tech lobbyists, investors, and former Trump advisers who warn the administration is undermining its own deregulatory, export-driven AI agenda. Politico reports observers describe the situation as "bitterly ironic" given Trump's stated goal of American AI dominance.
House Committee Advances Child Safety Bills Despite Constitutional and Efficacy Concerns
The House Energy & Commerce Committee advanced a package of bills aimed at children's online safety, despite significant opposition from civil liberties groups. According to Bloomberg, the measures inject fresh momentum for a "light-touch" US approach as other countries implement stricter protections. However, CDT's analysis warns that many of the privacy protections are "neither strong nor constitutional" and may not survive First Amendment scrutiny.
CDT's op-ed in Tech Policy Press argues that families themselves suggest the bills won't work as intended, citing research showing parents want different approaches than age verification mandates. The bills appear designed to avoid the political pitfalls that stalled previous federal legislation, but CDT's formal letter to the Committee outlines specific free expression risks. The disconnect between legislative intent and practical implementation echoes similar issues that have plagued state-level efforts in this area.
UK Delays AI Copyright Decisions Amid Creative Industry Pressure, While Peers Warn Against Speculative AI Gains
The UK government is returning to the drawing board on AI copyright rules after proposals to let tech firms use creative works without permission triggered intense backlash from novelists, artists, and writers. According to The Guardian, a House of Lords committee has urged ministers to develop a licensing regime instead and abandon proposals for blanket AI access to copyrighted material. The committee specifically warned that the UK's creative industries "must not be sacrificed in the pursuit of speculative gains in AI technology." The Financial Times reports the government is delaying difficult decisions as it prepares to reveal the economic cost of the original proposals.
This represents a significant policy retreat for the UK government, which had positioned AI copyright reform as central to its technology strategy. The House of Lords intervention is particularly significant given peers' role in detailed policy scrutiny and their ability to delay legislation. The creative industries' mobilization has proven more effective than anticipated, forcing ministers to reconsider the balance between AI innovation and IP protection. Other jurisdictions, including the EU, are watching closely as the UK's approach could influence international norms.
US Drafts Sweeping Chip Export Controls Tying Global Sales to Investment Pledges
The Trump administration is considering sweeping new semiconductor export controls that would require countries to invest in the United States in exchange for access to advanced AI chips from Nvidia and AMD, according to Bloomberg and The Financial Times. Under the draft proposal, the US government would require permits for chip exports regardless of destination country, fundamentally reshaping global semiconductor trade. TechCrunch reports the government would play a role in every chip export sale, representing unprecedented extraterritorial reach.
The proposal would effectively weaponize US dominance in AI chip design to extract foreign investment commitments, combining export controls with industrial policy in a novel way. This goes far beyond current controls targeting China and other adversaries, potentially requiring allied nations to negotiate investment deals to maintain chip access. The semiconductor industry has not yet publicly responded to the draft, though previous export expansions have prompted concerns about competitiveness and supply chain disruption. If implemented, this would represent the most aggressive use of US export control authority in the technology sector's history.
Signals & Trends
Government AI Procurement Becoming Battleground Over Acceptable Use Policies
The Anthropic-Pentagon standoff reveals a fundamental tension: corporations building general-purpose AI models want usage restrictions to manage liability and reputation, while government agencies demand unrestricted access for national security applications. This isn't just about one contract—it's a template for future conflicts. The Pentagon's unprecedented designation of a US company as a supply-chain risk establishes government willingness to use procurement leverage to override corporate governance decisions. Meanwhile, OpenAI's immediate positioning to capture Anthropic's lost business demonstrates how government demands will shape competitive dynamics in the AI industry, potentially rewarding companies with fewer ethical guardrails.
Child Safety Legislation Advancing Despite Implementation Gaps and Constitutional Questions
Multiple jurisdictions are pushing forward with child online safety measures despite recurring concerns about effectiveness and constitutionality. The US House package advances with acknowledged constitutional vulnerabilities, India's Karnataka state and Indonesia are implementing social media age bans, and Europe is convening special panels on child safety. The pattern suggests political pressure to "do something" is overwhelming technical and legal objections. However, the consistent criticism from civil liberties groups and the lack of evidence that similar measures work indicates an implementation gap between legislation and actual child protection outcomes. This creates regulatory uncertainty for platforms and potential First Amendment litigation in the US.
Export Controls Evolving From Security Tool to Economic Leverage Mechanism
The draft US proposal to tie chip exports to foreign investment pledges represents a significant expansion of export control philosophy. Traditionally used to deny adversaries access to strategic technologies, controls would now extract economic concessions from allies. Combined with ongoing chip restrictions on China and the Anthropic designation, a pattern emerges: the US government is testing how far it can push extraterritorial authority in technology governance. This risks fragmenting global technology markets and accelerating other nations' efforts toward semiconductor independence, potentially undermining the long-term effectiveness of US export controls as China, Europe, and others invest heavily in domestic capabilities.
Explore Other Categories
Read detailed analysis in other strategic domains