Peter Steinberger, the Austrian programmer who created OpenClaw and was hired by OpenAI in February, declared 2026 “the year of the general agent” in an AFP interview published Monday during ClawCon in Tokyo. The interview is Steinberger’s most substantive public statement since OpenAI CEO Sam Altman announced the hire, saying Steinberger would “drive the next generation of personal agents.”

“2023-2024 was the year of ChatGPT, last year was the year of the coding agent, this year’s going to be the year of the general agent,” Steinberger told AFP.

OpenClaw Couldn’t Have Come From Big Tech

Steinberger was blunt about why OpenClaw originated from a solo developer rather than a major AI lab: the incumbents would have killed it in review.

“What you have to know about OpenClaw is, like, it couldn’t have come from those big companies,” he said. “Those companies would have worried too much about what could go wrong instead of just, like — I wanted to just show people I’ve been into the future.”

He built OpenClaw in November 2025 while experimenting with AI coding tools to organize his personal digital life, as previously reported by Bloomberg. The tool can now autonomously operate apps, web browsers, and smart home appliances through messaging apps like WhatsApp and Telegram.

Steinberger said the next breakthrough in AI agents will likely come from “someone who just wants to have fun” rather than a corporate lab.

Security Concerns and the ‘Cottage Industry’ Problem

Steinberger acknowledged worrying about misuse. His concern centers not on OpenClaw itself, but on third-party companies lowering the barrier to entry without educating users.

“Yes, I do worry a bit, especially because there’s now a whole cottage industry of companies that try to make a big buck and make it even simpler to install OpenClaw,” he told AFP.

He said he deliberately chose not to simplify the installation process: “I purposefully didn’t make it simpler so people would stop and read and understand: what is AI, that AI can make mistakes, what is prompt injection — some basics that you really should understand when you use that technology.”

When asked whether the risks meant tools like OpenClaw shouldn’t be built at all, he used an analogy: “If you build a hammer… you can hurt yourself. So should we not build hammers any more?”

China’s Lead and Moltbook’s Weirdness

On the US-China adoption gap, Steinberger told AFP: “If you see it as a competition, it certainly looks like China is gaining a lot of momentum.” He qualified the statement by noting “there’s still quite a bit of a leap between the best models from China and the best models in the US.”

This echoes his Bloomberg interview from last week, where he described a sharper version of the same divide: in some US companies, using OpenClaw could get you fired; in many Chinese companies, not using it could.

On Moltbook, the Reddit-like social network where OpenClaw agents converse autonomously, Steinberger pushed back on the narrative that AI had developed its own social dynamics unprompted. “A lot of that was, in my view, very much driven by humans to just create those stories,” he said.

ClawCon Tokyo

The interview took place at ClawCon in Tokyo, where hundreds of attendees — many dressed as lobsters — watched live OpenClaw demos and received help installing agents. Steinberger said joining OpenAI has given him “more resources to use on cool ideas.”

“I love that I helped a lot of people to bring AI from this scary thing into something that is fun and weird and gets them excited, because we need to make it good for this next century,” he said. “We need more people to think more about AI.”