Who Gets Access to the Machine
SandboxAQ has teamed up with Anthropic to integrate its scientific AI models directly into Claude, putting drug discovery and materials science tools behind a conversational interface that requires no specialized computing infrastructure to use. The setup lowers the technical barrier, but the people most likely to benefit are still the ones already inside large corporate and research institutions, where the apparatus of science is tied to profit and scale.
The company says its large quantitative models, or LQMs, are “physics-grounded” and are built on the rules of the physical world rather than patterns in text. SandboxAQ says the models can run quantum chemistry calculations and simulate both molecular dynamics and microkinetics, the study of how chemical reactions unfold at the molecular level. In the company’s own framing, these are not just research tools but “AI models engineered for the quantitative economy, a $50+ trillion sector spanning biopharma, financial services, energy, and advanced materials.”
The Corporate Pipeline
Founded roughly five years ago as an Alphabet spinout, SandboxAQ counts Eric Schmidt, Google’s former CEO, as its chairman. The company has raised more than $950 million from investors and has built out a number of different business lines, including a cybersecurity business. The money, the boardroom pedigree, and the expansion into multiple sectors all point to the same arrangement: capital funds the tools, and the tools are built to serve capital.
Nadia Harhen, SandboxAQ’s general manager of AI simulation, said, “For the first time, we have a frontier [quantitative] model on a frontier LLM that someone can access in natural language.” That access, however, is not the same as open access. Harhen said previously users of SandboxAQ’s LQMs would have had to provide their own digital infrastructure to run the models. The new interface removes one obstacle, but not the larger hierarchy deciding who gets to use the system and for what ends.
Harhen said SandboxAQ’s customers tend to be computational scientists, research scientists, or experimentalists, generally people who work at large pharmaceutical or industrial companies and are searching for new materials that can become marketable products. That customer base makes the target clear: not communities, not public labs, not mutual aid science, but firms with the resources to turn research into commodities.
What the Company Says It Solves
Harhen said, “Our customers come to us because they’ve tried all the other software out there, and the complexity of their problem is such that it didn’t work or didn’t yield positive results for them when that translation went to take place in the real world.” The quote lays out the logic of the system in plain language: the problem is not just scientific difficulty, but the failure of existing software to produce usable results for the market.
The article says the company is focused on who can actually use the science, rather than only on the science itself. That emphasis on usability is not neutral. It means the science is being packaged for those already positioned to extract value from it, while the rest are left outside the gate, watching another layer of high-tech power get folded into the hands of corporate actors.
The integration with Claude is presented as a breakthrough in accessibility, but the structure underneath remains the same: a heavily funded Alphabet spinout, chaired by a former Google CEO, selling advanced scientific tools into the “quantitative economy.” The interface may be conversational. The hierarchy is not.