Anthropic’s Call for an AI Slowdown Meets the Momentum of a Global Arms Race

Anthropic, now valued near $1 trillion, issued a striking appeal this week: the world should consider slowing or even temporarily halting frontier AI development. In a blog post for the Anthropic Institute, Marina Favaro and Jack Clark warned that internal advances are pushing models toward recursive self-improvement—where AI could enhance itself without human intervention. They proposed global agreements and verification methods modeled on nuclear treaties, acknowledging that monitoring training runs is far harder than tracking missile silos.

But the suggestion lands in a crowded arena. Rivals race ahead. Governments push national advantage. And history shows coordination at this scale rarely sticks. TechRadar laid out the obstacles: a meaningful pause would require multiple well-resourced labs across countries to agree on identical terms and verify compliance. “The incentive to defect quietly is enormous,” Anthropic’s own analysis noted, “because whoever continues while others pause could inherit the lead.”

The timing feels deliberate. Anthropic disclosed fresh data: engineers now merge eight times as much code per quarter compared with 2021-2025 levels. More than 80 percent of code merged into its codebase in May 2026 came from Claude. Success rates on open-ended tasks climbed 50 percentage points in six months. On benchmarks like SWE-bench, models moved from low single digits to saturation in two years. These numbers paint acceleration.

Dario Amodei, Anthropic’s CEO, has walked a consistent line. In his essay “The Adolescence of Technology,” he outlined both enormous opportunity and serious risk—from AI enabling a “country of geniuses in a datacenter” to potential mass displacement of white-collar jobs within one to five years. Yet he also sketched utopian outcomes: compressed decades of biological progress, cures for most disease, doubled human lifespans. “I believe if we act decisively and carefully, the risks can be overcome,” he wrote.

Critics pounced. David Sacks labeled the move regulatory capture. Ethan Mollick described a mix of marketing, lawyers and true believers inside frontier labs. The Wall Street Journal noted Anthropic’s annualized run rate heading toward $50 billion. The contradiction sits at the heart of frontier AI firms today: safety calls from billion-dollar players who keep shipping ever-stronger models.

China looms over every discussion. Amodei has stressed that the United States cannot afford to fall behind. Chip export controls remain the sharpest tool, he argues, because they create a genuine bottleneck. Yet Beijing builds its own stack. DeepSeek, Huawei and state-backed efforts close gaps. Real enforcement looks distant.

Short of enforceable treaties with teeth, the trajectory holds. AI will keep building on AI. Companies will keep shipping. Governments will keep competing. The option to slow down exists in theory. Practice tells another story.