OpenAI is making a significant bet on physical robotics after years focused on language models, with CEO Sam Altman announcing plans to develop humanoid robots capable of handling real-world tasks starting in 2027. The company is rebuilding a robotics division it had previously shelved, establishing new labs in California and launching an aggressive hiring campaign across hardware engineering, machine learning, and systems design.

Why Is OpenAI Returning to Robotics Now?

OpenAI's robotics journey isn't new. The company previously developed Dactyl, a robotic hand with impressive dexterity, but shut down that division around 2020 to focus resources on language models, the bet that eventually produced ChatGPT. The quiet rebuild began in February 2025, when OpenAI started reassembling its robotics capabilities.

The timing reflects a shift in what's now possible. Altman has linked this strategic pivot to advancements in embodied intelligence and simulation technology, essentially training AI systems to understand and predict physical environments. This represents a meaningful evolution from the earlier Dactyl era, when the focus was narrower and more task-specific.

What Are OpenAI's Robotics Plans for the Next Few Years?

Altman laid out a two-phase vision for the company's robotics ambitions. The first phase targets infrastructure and construction, with robots designed to assist skilled workers on job sites. The second phase aims at personal robots built for everyday consumer use. This phased approach is pragmatic: construction sites are controlled environments with defined tasks, high labor costs, and chronic worker shortages. A robot that can assist an electrician or carry materials doesn't need to navigate the unpredictable chaos of a family kitchen.

The company's infrastructure includes a new lab in San Francisco that has reportedly quadrupled in size since its inception, with plans to expand to a second facility in Richmond, California. Altman has signaled an aggressive hiring campaign across multiple disciplines.

How Is OpenAI Positioning Itself in the Robotics Race?

• Competitive Landscape: OpenAI is entering a robotics market that already has formidable competitors, including Tesla's Optimus humanoid project, which has been progressing through iterative prototypes
• Strategic Investments: OpenAI holds investment stakes in Figure and 1X, two humanoid robotics startups that have attracted significant venture capital, giving the company exposure to multiple approaches
• Market Opportunity: The broader humanoid robotics market has attracted projections ranging into multi-trillion-dollar territory over the coming decades, making this a high-stakes bet for the company

The decision to target construction before consumer applications reflects a realistic understanding of where robots can add immediate value. Infrastructure and construction represent a more forgiving testing ground than the unpredictable home environment.

What Other Challenges Is OpenAI Addressing?

Beyond robotics, OpenAI is also mobilizing the broader tech industry around cybersecurity. More than 100 tech companies, including Anthropic, Amazon Web Services (AWS), Google, and Microsoft, have joined OpenAI for what the company calls "collective action" on cyber defense.

"There is a limited window to strengthen cyber defences," stated Greg Brockman, OpenAI President and Co-Founder.

Greg Brockman, President and Co-Founder at OpenAI

The initiative reflects growing concerns that AI-powered cyberattacks will become increasingly sophisticated as AI models continue to grow more capable. According to OpenAI's open letter, the companies and public services communities depend on, from hospitals to water treatment plants to internet infrastructure, are at risk.

OpenAI outlined a collective approach for tackling this threat, calling for greater investment in security tools and resources, particularly for critical infrastructure. The company identified several vulnerabilities that leave systems exposed:

• Technical Weaknesses: Persistent software vulnerabilities, excessive user permissions, and configuration errors create entry points for attackers
• Outdated Systems: Unpatched systems and legacy technology lack modern security protections and are difficult to update
• Authentication Gaps: Weak authentication mechanisms fail to verify user identity and prevent unauthorized access
• Resource Constraints: Security teams, especially those responsible for critical infrastructure, have historically lacked sufficient resources and need greater access to tools and support

OpenAI advocates using AI itself to strengthen the capabilities of cybersecurity defenders, while encouraging organizations to share tools, knowledge, and verified fixes. The company called for greater cooperation between companies, governments, and security teams worldwide to raise cybersecurity standards and address emerging threats.

These dual initiatives, robotics and cybersecurity, reveal OpenAI's broader ambitions: moving beyond language models into physical systems while simultaneously addressing the security risks that advanced AI capabilities create. The robotics push signals confidence in the company's technical capabilities, while the cybersecurity coalition suggests recognition that AI's rapid advancement requires industry-wide coordination to manage emerging risks.