OpenClaw founder accepts Bloomberg interview: The U.S. can learn from China's adoption of AI for crayfish.

The popular open-source AI agent tool OpenClaw founder Peter Steinberger recently pointed out in an interview with Bloomberg reporter Shirin Ghaffary that the acceptance levels of AI agents in the U.S. and China have formed starkly contrasting corporate cultures.
(Background: OpenClaw will establish the “Lobster Foundation” for independent operations! Nvidia and ByteDance have confirmed their participation.)
(Additional context: OpenAI announced the closure of the Sora App, and Disney’s $1 billion collaboration project has collapsed: a major failure in social platform positioning.)

Table of Contents

Toggle

• China has become the world’s largest AI agent testing ground
• America’s conservatism versus China’s leap forward
• Steinberger joins OpenAI: Codex is the next step
• OpenClaw Foundation: Playing the role of “Switzerland”

In China, students, office workers, and even the elderly are testing OpenClaw, with some companies even mandating its use among employees. Although Chinese authorities have begun to restrict state-owned enterprises and government agencies from using it, China has evidently become a massive experimental ground for AI systems to take over people’s digital lives.

In the United States, while OpenClaw has been warmly welcomed by developers and early adopters, it has not sparked a comparable level of enthusiasm among the general public. Due to safety concerns about AI agents potentially “going rogue,” some American companies have begun to restrict employee usage.

OpenClaw founder Peter Steinberger recently accepted an exclusive interview with Bloomberg reporter Shirin Ghaffary, where he shared his observations on the disparity between the adoption of OpenClaw in the U.S. and China. Ghaffary noted during the interview that this is not just a difference in adoption speed, but rather a struggle between two completely different corporate logics. Bloomberg reported this in an article titled “OpenClaw Founder Says the U.S. Can Learn from China’s Adoption of AI.”

China has become the world’s largest AI agent testing ground

According to Bloomberg, the situation in China is almost like another planet. Students, workers, and the elderly are flocking to test OpenClaw; some Chinese companies not only encourage but also mandate employees to use it, with some companies even displaying spreadsheets listing each employee’s name and a column for “what was automated today”: KPI management has extended into the AI agent era.

Data shows that usage in China is nearly twice that of the United States, with Baidu planning to integrate OpenClaw into its mobile search service, which has about 700 million users. Alibaba has launched “JVS Claw,” ByteDance has rolled out “ArkClaw” on Volcano Engine, and Tencent Cloud is offering free installation plans in 17 cities.

Paradoxically, this wave has not made Beijing feel at ease. Ghaffary mentioned in the article that Chinese authorities have restricted the use of OpenClaw by state-owned enterprises and government agencies due to concerns about data flow and security boundaries of open-source tools. A country that is both the most active promoter of AI agents and the most sensitive regulator.

America’s conservatism versus China’s leap forward

In contrast, the scene in the United States is entirely different. Developers and early adopters are enthusiastic, but the mainstream market’s excitement is far from that of China. Some American companies have even begun to limit employee usage of OpenClaw under the guise of safety concerns.

This conservatism is not without reason. Steinberger expressed his deep concern to Bloomberg about a recent incident where a Meta security researcher was mocked: the researcher publicly pointed out the risks of misusing agent tools but was met with ridicule instead of serious attention.

Steinberger said, “If everyone is laughing, it will make people afraid to speak up.” He believes that the discussion space around AI agent safety issues should not shrink due to fear of being laughed at.

Steinberger joins OpenAI: Codex is the next step

Currently, Steinberger formally joined OpenAI on February 15, 2026, to lead the development of Codex agent technology. OpenAI CEO Sam Altman publicly referred to him as a “genius” and stated that he would drive the evolution of the next generation of personal AI agents.

Currently, Codex has over two million users weekly. Steinberger told Bloomberg that his vision is to gradually erase the line between “whether to code or not”: when AI agents can write and execute code on behalf of humans, programming will no longer be the privilege of a few but a foundational skill accessible to everyone.

Nvidia CEO Jensen Huang referred to OpenClaw as “the most important software release in history” at the Morgan Stanley TMT conference in early March, stating that “every company needs an OpenClaw strategy, just as they needed HTML and Linux strategies back in the day.”

OpenClaw Foundation: Playing the role of “Switzerland”

On the other hand, Steinberger is actively working on establishing the OpenClaw Foundation, which is expected to complete legal procedures in a few weeks. Nvidia has officially joined, ByteDance is also on board, and Tencent is set to follow suit; Steinberger has revealed that discussions have started with Microsoft.

He told Bloomberg that he hopes the OpenClaw Foundation will play the role of “Switzerland”: maintaining a neutral, open stance that does not align with any single camp amidst the larger framework of U.S.-China tech rivalry, allowing open-source AI agent tools to become a globally shared infrastructure rather than a pawn in geopolitical struggles.