Sam Altman says Noam Brown joining OpenAI after 10-year wait
Synopsis
OpenAI CEO Sam Altman announced on 18 June 2026 that AI researcher Noam Brown — known for the Libratus and Pluribus poker AIs — is joining OpenAI, a goal Altman says he has held since the company's founding a decade ago.
Key Takeaways
Sam Altman publicly announced on 18 June 2026 that Noam Brown is joining OpenAI .
Altman stated he had wanted to work with Brown 'since the very beginning of OpenAI,' placing the ambition at the company's December 2015 founding.
Noam Brown is best known for creating Libratus ( 2017 ) and Pluribus ( 2019 ), landmark AI systems that defeated professional poker players.
Brown's prior work at Carnegie Mellon University and Facebook AI Research established him as a leading figure in multi-agent reinforcement learning.
The hiring reflects intensifying competition among frontier AI labs for researchers with expertise in game theory and strategic reasoning.
Specific responsibilities and formal start date have not been publicly confirmed beyond Altman's post.
OpenAI chief executive Sam Altman announced on Thursday, 18 June 2026 that celebrated AI researcher Noam Brown is joining OpenAI, describing it as the fulfilment of a decade-long ambition. Altman posted on X that Brown is 'one of the people I have most wanted to work with since the very beginning of OpenAI,' adding that it 'only took 10 years' and expressing confidence the wait would prove worthwhile.
Context
Noam Brown is one of the most closely watched researchers in the field of multi-agent reinforcement learning. He rose to prominence with Libratus in 2017, a poker-playing AI developed at Carnegie Mellon University that defeated professional players, and followed it with Pluribus in 2019, the first AI to beat elite humans in six-player no-limit Texas Hold'em — a landmark result published while Brown was at Facebook AI Research (FAIR).
Both systems demonstrated that AI could master imperfect-information games, a class of problems directly relevant to real-world reasoning, negotiation, and strategic planning. The results were widely cited within OpenAI's own research lineage and made Brown a figure that frontier labs tracked closely.
Policy Backdrop
OpenAI was founded in December 2015 as a non-profit with a mission to develop safe artificial general intelligence. Altman's post frames Brown's arrival as something he had envisioned from that founding moment, suggesting the recruitment effort spans essentially the company's entire history.
The broader industry context is one of intense competition for a small cohort of researchers with proven records in game theory, multi-agent systems, and large-scale training. High-profile hiring moves have repeatedly reshaped the competitive landscape among frontier labs, and Brown's move to OpenAI is among the more significant such transitions in recent memory.
Stakeholders and Impact
For the AI research community, Brown's affiliation shift carries weight beyond symbolism. His expertise in multi-agent reasoning and strategic decision-making is directly applicable to the next generation of AI systems that must coordinate, negotiate, or plan under uncertainty — capabilities central to OpenAI's roadmap toward more capable models.
Rival organisations, particularly Meta and Google DeepMind, will note the departure of a researcher whose foundational work was conducted partly within their own ecosystems. Talent retention at this level of the field has become a strategic priority as labs race to build systems beyond current large language model benchmarks.
What's Next
Altman's unusually personal framing — 'I think it will be worth the wait' — signals that Brown is expected to contribute to work the company considers consequential, not merely supplementary. Observers will watch OpenAI's forthcoming research releases and internal model announcements for Brown's name in author lists or acknowledgements.
The move also raises questions about what projects at OpenAI are now drawing on multi-agent and game-theoretic approaches, areas that have seen renewed interest as labs push toward systems capable of longer-horizon reasoning and autonomous decision-making.
Point of View
Beyond standard language modelling. For rival labs, the departure of a researcher of Brown's stature is a reminder that the talent war at the frontier tier remains acute and unresolved. Altman's public enthusiasm also functions as a recruitment message in itself, reinforcing OpenAI's culture of openly celebrating the people it brings in.
NationPress
5 Aug 2026
Frequently Asked Questions
Who is Noam Brown and why is he important in AI?
Noam Brown is an AI researcher who created Libratus (2017) and Pluribus (2019), the first AI systems to defeat professional players in heads-up and six-player no-limit Texas Hold'em respectively. His work on multi-agent reinforcement learning and imperfect-information game theory is considered foundational in the field.
Why is Noam Brown joining OpenAI significant?
Brown is one of the most accomplished researchers in multi-agent AI, an area increasingly relevant to building AI systems capable of strategic reasoning and autonomous decision-making. OpenAI CEO Sam Altman said he had wanted to work with Brown since the company's founding in 2015, underscoring the hire's importance.
What did Sam Altman say about Noam Brown on X?
Altman posted that Brown is 'one of the people I have most wanted to work with since the very beginning of OpenAI,' noted it 'only took 10 years,' and said he believed the wait would be worth it.
Where did Noam Brown work before joining OpenAI?
Noam Brown conducted his landmark poker AI research at Carnegie Mellon University and later at Facebook AI Research (FAIR), where Pluribus was developed and published in 2019.
What will Noam Brown work on at OpenAI?
OpenAI has not publicly specified Brown's exact role or responsibilities. Given his background, researchers expect his work to touch on multi-agent systems, strategic reasoning, and related capability research, though no official details have been confirmed beyond Altman's announcement post.
