GLM-52 897 —
GPT-56SC 873 —
CL-OP5X 865 -0.9%
GROK-46H 865 -0.9%
GEM-37FH 865 -0.9%
GPT-56T 861 —
GLM-5 856 —
MUSE-SPK 841 —
QWEN-38X 824 -2.3%
GPT-6A 820 —
KIMI-K3X 810 -1%
CL-FAB5H 787 -0.9%
CL-OP5H 764 -0.9%
CL-OP46H 742 -0.9%
CL-OP47H 733 -1.1%
GEM-38FH 676 -1%
CL-OP47 586 -0.5%
INKL 531 —
CL-OP46 497 —
CL-OP48 490 -0.2%
GLM-52 897 —
GPT-56SC 873 —
CL-OP5X 865 -0.9%
GROK-46H 865 -0.9%
GEM-37FH 865 -0.9%
GPT-56T 861 —
GLM-5 856 —
MUSE-SPK 841 —
QWEN-38X 824 -2.3%
GPT-6A 820 —
KIMI-K3X 810 -1%
CL-FAB5H 787 -0.9%
CL-OP5H 764 -0.9%
CL-OP46H 742 -0.9%
CL-OP47H 733 -1.1%
GEM-38FH 676 -1%
CL-OP47 586 -0.5%
INKL 531 —
CL-OP46 497 —
CL-OP48 490 -0.2%
← Back to feed

1,100 AI Lab Employees Sign 'Pacing the Frontier' Letter Asking U.S. to Build International Pause Tools

More than 1,100 employees across the four largest frontier AI labs signed an open letter on July 28 asking the U.S. government to invest in tools that could slow AI development if risks escalate. The letter, published at pacingthefrontier.com with organizational backing from Guidelight AI Standards and Encode AI, does not call for an immediate pause — it requests that government develop the option to pace, should one become necessary.

“To realize AI’s potential, industry, government, and society at large may need the option to buy time to address emerging risks, develop security measures, and strengthen oversight,” the letter states.

What the Letter Actually Asks For

The petition is explicitly not a moratorium. Signatories are asking the U.S. to pursue an international effort to develop:

  • Technical pacing tools: mechanisms that could modulate training runs or deployment velocity at the frontier
  • Governance infrastructure: international coordination frameworks that don’t require unilateral U.S. action
  • Security measures: hardening of frontier systems against theft, misuse, and undetected capability gains

The framing is defensive, not prohibitive. The signatories argue that without pacing infrastructure, governments will have no tools to respond if a future model demonstrates unexpected capability jumps.

The Signatory Base

Employees from OpenAI, Anthropic, Google, and Meta are named as signatories. At over 1,100 names, this represents a meaningful slice of the research and engineering staff at the labs closest to the frontier — including at labs whose official positions range from supportive of safety measures (Anthropic) to skeptical of formal regulation (xAI, which is notably absent from the named labs).

The timing follows a sequence of capability milestones that have accelerated the internal safety debate: GPT-5.6 Sol’s documented ability to self-replicate attack chains, Fable 5’s 93.9% SWE-bench score, and the government’s own gating of GPT-5.6 Sol access for 13 days this month.

What It Doesn’t Do

The letter has no mechanism. It cannot compel action from any government, and the labs themselves retain control over training and deployment schedules. Several signatories likely work on the very models they’re asking to keep regulated.

What the letter does accomplish is establishing political cover: if regulators move toward international AI governance frameworks in the next 12 months, proponents can point to 1,100+ industry employees who publicly supported the premise.

The call for technical pacing infrastructure — as opposed to legislation alone — is the most substantive element. It signals that at least a portion of the technical community believes the tools to pace AI development don’t yet exist, and that building them should be a government priority before they’re needed.