The jump of Peter Steinberger OpenAI has become one of the most talked-about moves in the artificial intelligence sector. The creator of OpenClaw, one of the most influential open agents of the moment, joins the laboratory responsible for ChatGPT in the midst of a global battle for talent specializing in agentic AI.
Far from meaning the project's closure, the agreement implies that OpenClaw will be reorganized as a foundation Open source, backed by OpenAI but with an explicit commitment to remaining independent and community-focused. The move comes amid fierce competition among giants like Google, Anthropic, xAI, and Meta to lead the next generation of intelligent agents.
Who is Peter Steinberger and how was OpenClaw born?
Steinberger, an Austrian-born software developer, was already known in the tech ecosystem for previous projects such as PSPDFKit, a company focused on PDF processing. After years dedicated to that business venture, he had partially stepped away from the spotlight, until he decided to experiment with AI agents, initially as an almost artisanal project.
That experiment mutated in name and form: first Clawdbot, later Moltbot and finally OpenClawThe tool gained international recognition under this name. It became popular due to its practical approach: an agent capable of perform real actions instead of simply answering questions, something that fit very well with startups and technical teams looking to automate tasks without setting up complex infrastructures.
According to Steinberger himself, the growth was much faster than expected. In just a few weeks, OpenClaw accumulated more than 100.000 stars on GitHub and it came to register around two million visits in just one week, figures that placed it among the most followed open source projects of the moment.
This sudden success turned what he himself defined as a "fun project" into a global phenomenonThe developer describes this period as a "whirlwind" in which proposals, investment suggestions, collaboration offers, and all kinds of pressure to turn OpenClaw into a large, traditional company began to arrive.
What OpenClaw actually does and why it matters
OpenClaw presents itself as a open-source AI agent It can take control of a large part of an individual's or company's digital workflow. It can manage email, coordinate calendars, check in for flights, deal with insurance companies, book trips, or interact with messaging apps without constant supervision.
In its most advanced version, it is even capable of control most of a computer's functionsacting as an assistant that executes complex tasks based on relatively simple instructions. This ability to "move" through the user's digital environment places it in a different category from that of purely conversational chatbots.
The design philosophy focuses on action: the goal is not just to respond, but solve real workThat's why many tech startups have looked to OpenClaw as an alternative for automate internal processes without deploying heavy enterprise solutions or hiring large development teams.
In parallel, its open infrastructure has allowed other projects to build upon its code. A striking example is MoltbookOpenClaw, a social network created specifically for autonomous agents, was designed by entrepreneur Matt Schlicht. On this platform, agents generated from OpenClaw interact with each other, even organizing "religions" or collective plots that have attracted the attention of the technology press.
Outside the West, the agent has also found traction. In ChinaFor example, OpenClaw has been combined with local models such as DeepSeek and it has begun to integrate with the country's messaging applications through custom configurations. Even Baidu It has explored offering direct access to OpenClaw from its main smartphone app, reflecting interest in this agent-based approach in Asia as well.
The future structure: an open-source foundation backed by OpenAI
One of the key points of the announcement is that, despite Steinberger joining OpenAI, OpenClaw will not be absorbed as a closed productBoth the developer and Sam Altman have insisted that the project will be transformed into a independent foundation, with the aim of preserving its open character and the freedom of evolution of the community.
Steinberger has repeatedly stated that for him it was essential that OpenClaw remain open source and maintain space for “thinkers, hackers, and people who want to control their data.” The foundation aspires to become that environment, accommodating new models and companies that want to integrate without being trapped in a closed, proprietary ecosystem.
OpenAI, for its part, is committed to providing technical support and some financial backing, in addition to remaining a sponsor of the project. The idea is to offer stability and resources so that the tool can continue to grow without abandoning its community-driven and experimental DNA.
This framework is relevant for Europe and Spain, where institutions and companies often pay close attention to issues such as governance, transparency and data controlA powerful agent like OpenClaw, managed by a foundation and with auditable code, is a better fit for regulatory requirements, including the future implementation of AI Act European.
In his public statements, Steinberger acknowledges that OpenClaw could have become a large, traditional company, but asserts that this path “doesn’t excite him.” His stated priority is “changing the world” and reaching as many people as possible, something which, in his view, requires an open model supported by an organization with resources and global reach like OpenAI.
OpenAI's vision: an extremely multi-agent era
Sam Altman, CEO of OpenAI, announced Steinberger's arrival at the social network X, emphasizing that his role will be to drive the next generation of personal agentsIn his words, the creator of OpenClaw is “a genius” with powerful ideas about “very intelligent agents that interact with each other to do very useful things for people.”
Altman also emphasized that OpenAI expects that approach to soon become a core part of its product offeringIn other words, agents capable of acting autonomously, in a coordinated and secure manner, would be one of the cornerstones of the company's future strategy, beyond basic models like GPT.
The executive has long advocated for the idea of a future extremely multi-agentIn this model, there isn't a single AI that does everything, but rather a coordinated ecosystem of specialized agents. Each agent would assume specific tasks, communicate with others, and adapt to the context and objectives defined by the end user.
With this move, OpenAI reinforces its commitment to this vision, integrating a figure who has demonstrated the ability to take the concept of an agent beyond theory and laboratory demonstrations. Steinberger's addition to the team related to Codex And with the integration of agents in ChatGPT, it opens the door for millions of users to interact with more advanced agent solutions in their daily lives.
For the European tech sector, this trend serves as a warning of where the dominant platforms in the market are heading: less focus on the simple chatbot and more on deep workflow automationThis is especially relevant for SMEs and large companies in Spain that are exploring AI for administrative tasks, customer service, or document management.
Risks, security and debates surrounding OpenClaw
The other side of the coin for an agent with such a capacity for action is risk. The very nature of OpenClaw—capable of controlling critical computer functions and accessing multiple services—has raised concerns. cybersecurity concerns between researchers and authorities.
In China, for example, the Ministry of Industry has warned that the agent could represent significant risks If configured improperly, it opens the door to cyberattacks, data leaks, or malicious use through extensions that conceal malwareResearchers in other countries have also pointed out the need to strengthen transparency about exactly what the agent does and how their behavior is monitored.
Steinberger is aware of these concerns and has pointed out that to “get agents everywhere” a much broader change in the way security is approached. Their reasoning is that, as agents gain autonomy, the responsibility for designing controls, permissions, and audits grows exponentially.
The alliance with OpenAI is presented, in part, as a way to accelerate that security agenda, leveraging access to state-of-the-art models and unpublished research which are not always within reach of community projects. During a recent stay in San Francisco, the developer held meetings with several leading laboratories, which helped him gauge the potential of working with a global organization.
For Europe, where AI regulation is moving towards a more stringent framework, these debates are significant. Companies and public administrations evaluating the use of agents like OpenClaw in Spain will have to weigh the balance between productivity and risk, and will probably require additional guarantees on issues such as stock recording, traceability, access limits and personal data protection.
A key move in the global war for AI talent
Steinberger's arrival at OpenAI comes in a context of talent in artificial intelligenceThe major players in the sector —Google, Anthropic, xAI, Meta or OpenAI itself— compete not only with products, but also to recruit the key minds that shape the technological agenda.
The figures involved in this race are enormous. OpenAI has made deals worth billions of dollars in recent months to strengthen its position, such as the purchase of a start-up linked to Jony Ive, known for his time at Apple. Meta, for its part, has closed multi-million dollar deals to acquire significant stakes in companies like Scale AI, with the aim of securing talent and technology.
Meanwhile, OpenAI is negotiating a new funding round which could be around $100.000 billion, with a potential valuation of several hundred billion. At the same time, direct rivals such as anthropic They have closed rounds of tens of billions to boost Claude, their alternative to ChatGPT.
In this context, adding the person behind one of the most successful agentive projects of the last year sends a clear message to the market: OpenAI not only wants to lead in foundational models, but also in Agent-based practical applicationsFor the European ecosystem, which often sees local talent migrate to large US labs, the movement reopens the debate on how to retain and support developers who achieve global traction.
Steinberger, who had already “played the game” of building and running his own company for over a decade, is now opting to integrate into a larger structure to accelerate his vision. His case illustrates a recurring tension in open-source projects: growing as startup either become independent or integrate into a giant that offers resources, at the cost of relinquishing some direct control.
Implications for Spain and Europe: regulation, business and adoption
Beyond the global headlines, the movement has concrete implications for Europe and, in particular, for countries like Spainwhere the digitalization of businesses and public administrations coexists with increasing regulatory demands. The combination of open agents supervised by a foundation This may be especially interesting for public and private projects seeking transparency and auditability.
In the business world, European companies are beginning to move from experiments with chatbots to real process automationTools like OpenClaw, integrated with leading models through OpenAI, can simplify tasks in sectors such as financial services, insurance, logistics or public administration, provided they are designed with the appropriate safeguards.
Furthermore, Steinberger's "agents for all" philosophy aligns with the European agenda of digital accessibilityHis aspiration to build an agent that even his own mother could use serves as a metaphor for tools designed for people without a technical background, something relevant in a continent with a diverse and aging population.
However, mass implementation in Europe will be conditional upon compliance with regulations on privacy, data protection and ethical use of AI. The fact that OpenClaw keeps its code open source and is overseen by an independent foundation can facilitate external audits, something that regulators and large European clients often value.
For the Spanish entrepreneurial ecosystem, the OpenClaw story offers several lessons: the potential of projects open source as a means of market validation, the importance of building tools that perform tasks and not just converse, and the role of the community as an engine of adoption and opportunities, even on a global scale.
With Peter Steinberger joining OpenAI and OpenClaw becoming a foundation, the agentic AI sector is entering a phase where applications capable of acting on our behalf are gaining prominence over simple conversational assistants. The combination of talent, open source, and the backing of an AI giant positions OpenClaw as a project to watch, particularly in Europe, where intelligent automation and responsible regulation are increasingly integrated into the technological and economic agenda.
