Humans&
Last reviewed
Jun 2, 2026
Sources
5 citations
Review status
Source-backed
Revision
v1 · 1,409 words
Improve this article
Add missing citations, update stale details, or suggest a clearer explanation.
Suggest edit
CtrlK
Last reviewed
Jun 2, 2026
Sources
5 citations
Review status
Source-backed
Revision
v1 · 1,409 words
Add missing citations, update stale details, or suggest a clearer explanation.
Humans& is an American artificial intelligence company founded in 2025 that describes itself as a "human-centric" frontier AI lab. The startup, whose name is stylized with a trailing ampersand, was established by researchers who had previously worked at Anthropic, xAI, Google, OpenAI and Meta. In January 2026 it announced a $480 million seed round at a $4.48 billion valuation, one of the largest seed financings on record. Rather than building autonomous agents intended to replace human labor, Humans& states that it is building foundation models aimed at social intelligence and at helping people coordinate and work together. [1][2]
Humans& is a frontier AI research company headquartered in the United States. It was incorporated in September 2025 and emerged from stealth on January 20, 2026, when it disclosed its seed round. At that point the company was roughly three months old and had not released a product. [1][3]
The company positions itself in contrast to the dominant strategy among large AI laboratories, which has centered on building increasingly capable autonomous agents to automate work. Humans& frames its thesis around coordination rather than conversation or automation: it argues that the next major frontier for AI lies in helping groups of people communicate, build trust, and act together, with AI serving as connective tissue inside organizations and communities rather than as a replacement for the people in them. [1][4]
The table below summarizes the company's key facts as reported at launch.
| Attribute | Detail |
|---|---|
| Name | Humans& |
| Type | Private artificial intelligence company |
| Founded | September 2025 |
| Announced | January 20, 2026 |
| Headquarters | United States |
| Seed funding | $480 million |
| Valuation | $4.48 billion |
| Lead investors | SV Angel; co-founder Georges Harik |
| Focus | Human-centric AI for collaboration and coordination |
Humans& was founded by a group of researchers drawn from several of the leading AI laboratories. Press coverage of the launch named five co-founders. [1][2]
Eric Zelikman is the co-founder and chief executive officer. He was an early employee at xAI, where he contributed to pretraining data for Grok 2 and worked on reinforcement learning for reasoning in later Grok models. Before joining xAI he was a Ph.D. candidate at Stanford University, where he authored the paper introducing STaR ("Self-Taught Reasoner"), an early method for training language models to reason using their own generated rationales. His doctoral advisers included Noah Goodman, a fellow co-founder of Humans&. [4][5]
Andi Peng is a co-founder and former Anthropic researcher. At Anthropic she worked on reinforcement learning and on the post-training of Claude models spanning the 3.5 through 4.5 generations. [1][2]
Yuchen He is a co-founder and a former xAI researcher who helped develop the Grok chatbot. He has described the company's plan to train its model in a way that has humans and AI systems interacting and collaborating together, using long-horizon and multi-agent reinforcement learning. [1][4]
Georges Harik is a co-founder and was Google's seventh employee, where he helped build the company's first advertising systems. He later worked on research including probabilistic programming and the Quiet-STaR method. He also participated in the seed round as an investor and co-led it. [1][2]
Noah Goodman is a co-founder and a professor of psychology and computer science at Stanford University. He was a doctoral adviser to Zelikman during the latter's time at Stanford. [1][4]
On January 20, 2026, Humans& announced that it had raised $480 million in a seed round at a post-money valuation of $4.48 billion. The company said the majority of the capital would be spent on compute for training its models. [1][2]
The financing drew attention because of its scale relative to the company's stage. At three months old and with no released product, the round was widely described as one of the largest seed rounds in venture capital history. Crunchbase News and other outlets reported it as the second-largest seed on record, behind only Thinking Machines Lab. [2][3]
| Company | Seed amount | Valuation | Date |
|---|---|---|---|
| Thinking Machines Lab | $2 billion | $12 billion | July 2025 |
| Humans& | $480 million | $4.48 billion | January 2026 |
Thinking Machines Lab, founded by former OpenAI chief technology officer Mira Murati, raised a $2 billion seed round in 2025 that Crunchbase described as by far the largest in its dataset. Humans& has been grouped with this small set of so-called mega-seed rounds, in which experienced AI researchers raise unusually large amounts of early capital on the strength of their track records. [2][3]
Humans& describes its mission in terms of human connection and collaboration. The company argues that progress for people comes from understanding one another, building trust, making connections, and working together, and it states that AI can be reimagined to center on people and their relationships rather than on individual task automation. In its framing, AI at its best should act as connective tissue that strengthens organizations and communities. [1][4]
Technically, the company says it is building a foundation model architecture optimized for social intelligence rather than for information retrieval or code generation. Its stated research areas include long-horizon and multi-agent reinforcement learning, memory systems, and modeling user understanding, all aimed at enabling planning and coordination among groups of people and AI systems over extended periods. The company has said it is designing its product alongside the model, and reporting has described its ambition as a coordination layer for teams that could eventually overlap with collaboration tools such as group messaging and shared documents. [1][4]
The seed round was led by the venture firm SV Angel together with co-founder Georges Harik. A range of additional investors participated. [2]
| Investor | Type |
|---|---|
| SV Angel | Venture capital (lead) |
| Georges Harik | Co-founder (co-lead) |
| Nvidia | Strategic / corporate |
| Jeff Bezos | Individual |
| GV (Google Ventures) | Venture capital |
| Emerson Collective | Investment firm |
| Forerunner | Venture capital |
| S32 | Venture capital |
| DCVC | Venture capital |
| Human Capital | Venture capital |
| Felicis | Venture capital |
| CRV | Venture capital |
Emerson Collective is the firm founded by Laurene Powell Jobs. Reporting noted that the round also included participation from many other investors beyond those named. [2]
Humans& attracted notice for two reasons that reinforced each other: the size of its seed round and the pedigree of its founding team. The combination of a roughly half-billion-dollar raise, a multibillion-dollar valuation, and founders with experience building models such as Claude and Grok placed the company among the most heavily capitalized early-stage AI ventures, despite having no product at announcement. [1][2]
The launch was also discussed as a notable example of a contrarian product thesis within the AI industry. While many leading laboratories framed their work around autonomous agents that operate with little human involvement, Humans& publicly committed to the opposite premise, that AI should augment human collaboration rather than replace human labor. Coverage treated the company as a test of whether "human-centric" or coordination-focused AI could compete with the prevailing agent-centric approach. [1][4]
The Humans& financing came during a period of very large early-stage fundraising by AI startups led by well-known researchers. The most prominent precedent was Thinking Machines Lab, founded by Mira Murati, which raised a $2 billion seed in 2025. Other labs founded by departing researchers from major firms, including efforts focused on advanced or "superintelligent" systems, raised similarly outsized rounds during the same period. Humans& fit this broader pattern in which investor demand for frontier AI talent produced seed rounds far larger than historical norms. [2][3]