Safe Superintelligence Inc. (SSI), the AI research lab Ilya Sutskever co-founded on June 19, 2024 alongside Daniel Gross (Apple’s former AI lead) and Daniel Levy, shortly after leaving OpenAI, announced a long-term strategic partnership with Nvidia on July 27, 2026. Under the deal, Nvidia is making a direct investment in SSI, and the two companies will collaborate on the technical development of Nvidia’s current and future compute platforms — with SSI gaining access to Nvidia’s next-generation Vera Rubin platform specifically, an arrangement SSI said would let it increase its available compute by an order of magnitude. The deal also marks a notable technical shift for SSI, which has reportedly been moving away from Google’s TPU chips toward Nvidia GPUs as its primary compute infrastructure.
The partnership is a significant, concrete milestone for a company that has otherwise operated almost entirely out of public view since its founding. SSI’s stated approach — deliberately avoiding commercial product releases and short-term revenue pressure in favor of what the company calls a “straight shot” toward safe, aligned superintelligence — is a genuinely unusual strategy in an AI industry where nearly every other major lab, including Sutskever’s former employer OpenAI, ships frequent public product updates and competes openly on benchmark performance and user growth.
A CEO Change Along the Way — and a Rebuffed Acquisition Attempt
SSI’s own leadership changed significantly before this Nvidia deal came together. Meta reportedly attempted to acquire SSI outright after the startup’s April 2025 funding round valued it at roughly $32 billion, an offer Sutskever turned down. Meta instead hired co-founder Daniel Gross directly, in July 2025, to help lead its own newly formed Meta Superintelligence Labs — at which point Sutskever stepped in as SSI’s CEO himself, a role Gross had originally held. Even after that departure and turned-down acquisition bid, SSI has continued operating with a genuinely small team, reported at somewhere between 20 and 50 employees depending on the source, without ever having published a research paper or shipped a public product — an unusually quiet profile for a company that had already reached a $32 billion valuation before this year’s Nvidia partnership.
Sutskever’s Own History With This Exact Debate
Sutskever’s departure from OpenAI is itself among the most closely documented episodes in recent AI industry history. As OpenAI’s chief scientist, he sat on the board that voted to remove Sam Altman as CEO in November 2023, before publicly stating he regretted his participation in that decision and signing the employee letter that called for the board’s own resignation and Altman’s reinstatement. Sutskever left OpenAI roughly six months later, in May 2024, to found SSI — a company whose entire premise, avoiding the exact kind of rapid, competitive product-shipping pressure that defines OpenAI’s own public strategy, reads as a direct extension of the tension that reportedly contributed to his departure in the first place, though Sutskever has not publicly characterized his motivations for founding SSI in those specific terms.
What This Means for Philippine Founders
SSI’s Nvidia partnership is a useful real-world data point for any Philippine founder weighing whether a slower, more deliberate product strategy can still attract serious institutional backing in a fast-moving industry — Sutskever has raised and now secured major strategic partnerships without shipping a single public product, a genuinely rare position that most startups, in the Philippines or anywhere else, will never be able to replicate given how much earlier-stage funding typically depends on visible traction. The more practical, broadly applicable lesson is Sutskever’s own credibility as a technical co-founder of OpenAI itself — it’s a reminder that in AI specifically, deep technical reputation among a small circle of researchers and major infrastructure providers can sometimes substitute for the conventional startup metrics (revenue, user growth, public launches) that most companies, including nearly every Philippine startup, still have to actually demonstrate to raise capital. Meta’s rebuffed acquisition attempt is itself worth noting: even a company as well-resourced as Meta was reportedly unable to simply buy SSI outright after Sutskever declined, settling instead for hiring away co-founder Daniel Gross directly — a reminder that a strong-willed founding team’s control over its own company’s direction can outlast even a highly motivated acquirer’s best offer, provided the founders are genuinely unwilling to sell. For a Philippine founder facing acquisition interest from a much larger company, that decision is a real, current example that turning down a serious offer — even one that would have provided a clean, well-funded exit — remains a genuine, defensible option when a founder believes the company’s independent mission is worth more than what’s on the table. Whether SSI’s own long-term bet ultimately pays off in the way Sutskever intends is still genuinely unknown — the company has, after all, never shipped a public product for the market to actually judge — but Nvidia’s willingness to invest directly in that unproven approach is itself a real, current signal that at least one of the industry’s most consequential infrastructure providers takes the underlying research bet seriously.
Share this article