Nvidia Takes Equity Stake in Ilya Sutskever Startup Safe Superintelligence

The artificial intelligence landscape is undergoing another structural realignment as hardware titan Nvidia announced a long-term strategic partnership with Safe Superintelligence Inc. on July 27, 2026. The agreement gives Nvidia an equity stake in the startup founded by Ilya Sutskever while securing rare access to its research to inform future chip design. For Nvidia, the alliance extends its circular financing model into its most structurally integrated form yet, coupling financial investment directly with hardware supply commitments.

Hardware Commitments and Research Access at Safe Superintelligence

Under the terms of the agreement, Nvidia will supply Safe Superintelligence with Vera Rubin compute systems, powering the startup’s rigorous research initiatives. In return, the chipmaker gains profound visibility into cutting-edge safety and scaling architectures developed by Ilya Sutskever’s research team. This direct feedback loop between hardware development and foundational model research allows Nvidia to tailor its next-generation silicon directly to the evolving bottlenecks of advanced artificial intelligence development.

Ilya Sutskever’s Departure From OpenAI and the Path to AGI

The launch of Safe Superintelligence follows Sutskever’s high-profile exit from OpenAI, where he served as chief scientist and co-founder. Announcing his departure after nearly a decade at the center of the generative AI revolution, Sutskever reflected on the journey that brought ChatGPT to global prominence. Le cofondateur d’OpenAI, Ilya Sutskever, a annoncé mardi en fin de journée qu’il quittait la start-up, star de la Silicon Valley depuis qu’elle a lancé ChatGPT et la révolution de l’intelligence artificielle (IA) générative.

Ilya Sutskever stated via Trends Le Vif that after almost ten years, he had made the decision to leave OpenAI.

Translated from the original announcement, Sutskever stated that after almost a decade, he made the decision to leave OpenAI. In his public messages, he expressed deep confidence in the organization’s trajectory, noting that the company’s trajectory has been nothing short of miraculous, and adding:

Ilya Sutskever stated via Trends Le Vif that he was convinced OpenAI would build artificial general intelligence that is both safe and beneficial.

That vision of artificial intelligence—systems featuring cognitive capabilities matching or exceeding human performance—remains the driving force behind his new venture. During an appearance at a TED conference on artificial intelligence in San Francisco, Sutskever offered a stark prediction regarding the trajectory of cognitive systems.

Ilya Sutskever stated via Trends Le Vif that a day would come when digital brains would become as good as, or even better than, our own.

Navigating Past Board Turmoil and Future Technology Horizons

Sutskever’s transition away from OpenAI follows a turbulent period in late 2023, when he sat on the board of directors that voted to remove co-founder Sam Altman. M. Sutskever, scientifique en chef d’OpenAI, siégeait au conseil d’administration qui a voté la révocation du directeur général et cofondateur Sam Altman en novembre dernier. Cette éviction surprise avait plongé la start-up basée à San Francisco dans la tourmente pendant quelques jours. Microsoft, principal investisseur d’OpenAI, avait rapidement proposé d’embaucher le patron déchu, tandis que la majorité du personnel menaçait de partir. M. Altman avait rapidement été réintégré. Ilya Sutskever had quitté le conseil d’administration, mais était resté à son poste chez OpenAI. Having departed to focus on his personal project, Sutskever bade farewell to his former colleagues.

Ilya Sutskever stated via Trends Le Vif that it had been an honor and a privilege to work together, and that he would miss them all.

OpenAI a présenté lundi une nouvelle version de ChatGPT qui peut désormais tenir des conversations orales et fluides avec ses utilisateurs, un pas de plus vers des assistants d’IA toujours plus personnels et performants. Google a de son côté dévoilé mardi ses dernières et nombreuses innovations en matière d’IA générative, lors de sa conférence annuelle pour les développeurs. Là encore, les progrès technologiques convergent vers des interlocuteurs virtuels omniscients et de plus en plus “intelligents”, censés simplifier la vie des utilisateurs.