Mark Filipino hosts this Financial Times news briefing covering major developments in Middle Eastern energy markets and AI military applications. The episode features Malcolm Moore, the FT's energy editor, discussing Iran's critical Karg Island oil terminal, and George Hammond, venture capital correspondent, analyzing Anthropic's lawsuit against the Pentagon.

The briefing examines why Karg Island remains untouched despite ongoing regional conflicts, explores the unprecedented classification of an AI company as a supply chain risk, and discusses broader implications for both global energy markets and military AI deployment.

Oil Markets Cool as G7 Prepares Emergency Reserve Release

Brent crude fell from $119 to below $90 per barrel after Trump declared the Iran war 'very complete, pretty much' and G7 ministers announced readiness to release emergency petroleum reserves

Countries implement energy conservation measures including Bangladesh closing universities, UK Automobile Association advising people to avoid driving, and Germany considering all options to contain rising energy prices

Karg Island: Iran's Untouchable Oil Export Lifeline

Karg Island, located 25 kilometers off Iran's coast, handles 'about nine out of every 10 barrels of oil that they load' due to Iran's shallow coastline preventing supertanker access elsewhere - Malcolm

Former US deputy special envoy Richard Nephew explains that 'the Iranian economy bottoms out without it' since Iran is an oil revenue-dependent economy

Iraq bombed Karg during the 1980s Iran-Iraq war, forcing shutdowns multiple times, but current conflicts have avoided targeting the facility despite attacks on other regional energy infrastructure

Attacking Karg would remove Iranian oil from global markets, potentially causing another price spike and feeding through to global economic instability

Saudi Arabia has threatened to 'blow up Karg Island' if Iran attacks their oil fields, while some suggest Marine seizure, though experts consider both scenarios extremely unlikely

Anthropic Sues Pentagon Over Supply Chain Risk Classification

Anthropic is 'currently the only AI company whose models are used in the most sensitive operations, so classified operations' but disputes Pentagon demands for 'all lawful purposes' usage - George

The company specifically opposes two AI applications: 'domestic mass surveillance' and 'the use of AI models for lethal autonomous weapons' during ongoing Pentagon negotiations

The Pentagon classified Anthropic as a supply chain risk, a designation 'normally reserved for companies from China or Russia' that forces military contractors to cut ties with the company

Anthropic claims the designation is 'arbitrary and capricious' and 'politically motivated' while seeking an injunction to allow continued military contract partnerships

The dispute highlights broader questions about AI governance, with 'OpenAI, and as we understand it, Google as well' facing similar debates about military AI applications