Capital & Industrial Strategy

The US Defense Department designated Palantir's Maven AI system as a 'program of record', according to Reuters and Bloomberg, marking a major escalation in the company's defence AI business. This classification converts Maven from a project into core military infrastructure, typically indicating multi-year funding commitments and integration into operational doctrine. At Palantir's developer conference, CEO Alex Karp doubled down on the company's vision of 'AI built for battlefield advantage', according to Wired, positioning the firm as the primary AI platform for US combat operations. The programme-of-record designation follows months of Palantir's aggressive positioning as a defence-first AI company at a time when competitors like Anthropic face Pentagon pushback over perceived national security risks.

Senate Democrats Elizabeth Warren and Richard Blumenthal are probing Nvidia's $20 billion licensing deal with AI inference startup Groq, questioning whether the structure improperly avoids merger review while illegally consolidating Nvidia's market power, reports Bloomberg. The deal — announced as a licensing arrangement rather than an acquisition — raises questions about whether companies are structuring transactions to circumvent CFIUS and DOJ scrutiny. Separately, CEO Jensen Huang projected $1 trillion in AI chip sales through 2027 at Nvidia's GTC conference, according to TechCrunch, underscoring the company's dominance in AI compute infrastructure. The Groq licensing model may become a template for how Nvidia extends its ecosystem without triggering regulatory review — a strategy that could reshape how AI market consolidation occurs.

US prosecutors charged Super Micro co-founder David Liaw, a Taiwan-based executive, and a contractor with conspiring to illegally divert Nvidia-powered servers to China, according to Reuters and Bloomberg. The indictment alleges Liaw bypassed export controls to ship restricted AI chips, a violation that carries severe penalties under US trade law. Super Micro's stock fell more than 33%, erasing a third of the company's market value, with Bloomberg reporting the company is now urgently shoring up compliance operations. Liaw resigned from Super Micro's board following the charges, according to Reuters. The case underscores how export control enforcement is tightening as AI compute becomes a national security priority, creating significant legal and reputational risk for hardware suppliers.

The White House unveiled a national AI policy framework explicitly designed to preempt state-level regulation, according to Reuters, Politico, and Bloomberg. The framework, delivered to a sceptical Congress, prioritises federal standards over state rules — a direct response to California and other states advancing their own AI liability and transparency laws. TechCrunch reports the plan shifts child safety responsibility toward parents rather than platforms, reflecting a lighter regulatory touch. The proposal arrives amid fierce intra-party debate, with Financial Times noting 'fierce Maga backlash' over federal overreach concerns. AI industry leaders have lobbied aggressively for federal preemption, arguing a patchwork of state laws would hobble US competitiveness, as reported by CNBC.

Halter, a New Zealand startup making AI-powered collars for cattle management, is in talks to raise funding at a valuation exceeding $2 billion — double its prior round — with backing from Peter Thiel's Founders Fund, according to Bloomberg. The deal highlights continued investor appetite for vertical-specific AI applications in sectors like agriculture, where AI adoption can deliver measurable productivity gains. The valuation — ambitious for a hardware-enabled AgTech play — reflects investor conviction that AI will reshape industries far beyond software, with tangible ROI in operational efficiency. However, Bloomberg separately quoted Accel partner Arun Mathew warning of 'bubble tendencies forming in AI', suggesting some late-stage valuations may not hold if market conditions tighten.