OpenClaw’s creator: in the U.S., using it will get you fired; in China, not using it will get you fired instead

The claim and the context

It has been reported that Peter Steinberger (彼得·斯坦伯格), the creator of the AI agent OpenClaw (“龙虾”), told Bloomberg that corporate responses to agent‑style AI in China and the United States could not be more different. In China, Steinberger said, employees are being pushed — and in some cases mandated — to adopt AI tools to drive productivity. In the U.S., by contrast, some firms are tightening restrictions because of safety and security worries.

Two rollout models collide

IT Home (IT之家) carried the interview on Phoenix (凤凰网), describing China as a vast live testbed where students, office workers and even older users swarm new agents. It has been reported that some Chinese companies actively require employees to use OpenClaw to explore efficiency gains. Rapid, top‑down adoption is familiar to Western readers who may not follow China’s tech ecosystem: large firms and local governments often push new digital tools quickly into daily workflows. But does speed come at a cost?

Caution and control in the U.S.

In the United States, developers reportedly celebrate OpenClaw, yet corporate policies have begun to curb its use. The reason is straightforward: concerns about autonomous agent behaviour, data leakage and regulatory exposure. Those fears sit alongside broader geopolitical pressures — export controls, sanctions and growing scrutiny of AI governance — that make U.S. companies more risk‑averse when it comes to experimental agents.

What Steinberger says next

Steinberger has since joined OpenAI to work on the Codex code‑generation team, it has been reported, and he argues the industry must move from tightly defined tools to fluid personal and workplace agents that can interoperate while protecting enterprise secrets and user privacy. The question he poses — and the one regulators and employers must answer — is whether rapid, enforced deployment or cautious, controlled adoption will better serve safety, innovation and workers’ rights. Which model will win?