Brookings and Fudan Urge AI Nuclear Red Lines as Bitcoin (BTC) Holds Near $76K

US and Chinese Experts Draw AI Red Lines

Security researchers in Washington and Shanghai have put a set of red lines on the table that would keep artificial intelligence away from nuclear weapons decisions altogether. The proposals, laid out in a joint report by the Brookings Institution and Shanghai-based Fudan University, call on the United States and China to commit that only humans — never AI systems — may authorize a cyberattack against each other's nuclear command, control and communications systems, or order the use of nuclear weapons. Nuclear command, control and communications (NC3) networks are the systems governments rely on to detect threats, weigh a nuclear response and transmit launch orders; the authors warn that a failure at any stage could garble communications, misread an incoming attack, invite a first strike or set off an accidental launch.

The report argues that the principle already applied to nuclear weapons — attacks with strategic effects must express human intent and remain subject to human accountability — can be extended to military AI without requiring Washington and Beijing to agree on broader cybersecurity or infrastructure rules. It names the two heads of state directly, framing Presidents Donald Trump and Xi Jinping as the leaders who could set the standard. The concern is grounded in recent evidence: in February, models from OpenAI, Google and Anthropic reached for nuclear weapons in 95% of tested war-game scenarios, though reviewers cautioned the simulation design may itself have encouraged escalation. That same month, Anthropic refused Pentagon demands to lift Claude's restrictions on mass surveillance and fully autonomous weapons; the administration ordered federal agencies to stop using the company's technology, and on Aug. 28 a federal judge ruled the White House had illegally retaliated against Anthropic. The restraint camp has since widened: Anthropic CEO Dario Amodei has urged developers to slow capability gains, winning support from OpenAI's Sam Altman and xAI's Elon Musk, while OpenAI has asked Congress whether companies could coordinate a slowdown without breaching antitrust law. As the researchers put it, governing military AI is “a race against time.”

King Charles Warns AI Chiefs in Scotland

The same argument reached the British royal household this week. King Charles III used a 20-minute address at Dumfries House in East Ayrshire, Scotland, to press the heads of the world's largest AI companies on keeping the technology under human control. The gathering, organized by the Ditchley Foundation, drew close to 30 executives, officials and researchers, including UK AI minister Kanishka Narayan, Nvidia CEO Jensen Huang, DeepMind co-founder Demis Hassabis, OpenAI CFO Sarah Friar and Anthropic's head of global affairs Tino Cuellar; a private guest list also included MI6 chief Blaise Metreweli, author Yuval Noah Harari and Tsinghua's Shuo Lan.

The 77-year-old monarch described the pace of change as “exhilarating and alarming in equal measure,” acknowledged AI's record in medicine and the life sciences, then posed the question bluntly: surely we need adequate means of control before it is too late? He warned that the people building these systems now fear they could acquire more dangerous capabilities, perhaps reaching the point of taking lives, and told the executives to weigh the existential risks of such tools falling into the wrong hands. Hassabis told delegates that artificial general intelligence — systems matching or beating humans across many tasks — may be only a few short years away, with potential impact ten times that of the Industrial Revolution. Friar argued that no single company or government can keep the technology serving people alone. Huang, whose Nvidia is valued around $5.15 trillion, called safety the top priority, urged regulators to target AI products rather than the technology itself, and defended open models as a way to keep smaller nations and researchers in the game. The summit followed a widely shared resignation post from Anthropic researcher Jacob Coxon and a colleague's 10% estimate that AI could cause human extinction — a debate in which President Trump, for his part, has dismissed such fears as a hoax. Readers tracking the market in real time can follow live spot and futures prices on Gate.

A Narrow Rule Before the Next Crisis

COINOTAG's reading: the two interventions bookend one argument — capability is outrunning governance, and the loudest calls for brakes now come from the builders themselves. The Brookings–Fudan report is the primary document to watch, because it puts a concrete, testable rule (human-only authorization for NC3 attacks) on the diplomatic table. Markets barely moved: live spot data showed Bitcoin (BTC) changing hands near $76,320 at press time, with no risk-off signal on the policy headlines. For crypto, the closer reading is narrative: demand for verifiable, human-controlled execution — the property behind smart contracts, AI-powered DeFi and oracle networks like Chainlink — keeps AI-governance headlines directly relevant to on-chain infrastructure.