Anthropic, the AI research and development company that produces the popular chatbot Claude, is deepening its roots in Cambridge. The San Francisco-based corporation announced on Tuesday that it will lease 24,000 square feet of office space at the One Kendall Square office complex for long-term use, according to a statement first obtained by The Boston Globe. The new space is large enough to accommodate 120 employees, including 50 of its existing Cambridge-based staff (currently located at a WeWork in Central Square).
“Cambridge is a natural place for Anthropic to put down roots,” Anthropic cofounder and president Daniela Amodei said in the statement. “It’s where some of the world’s leading life science researchers are using Claude to accelerate their work, among the more than a million people across Massachusetts who now use our products.”
Criticisms and concerns
The announcement was somewhat overshadowed, though, by posts made on X Tuesday night from former Anthropic employee Jacob Coxon, who said that AI’s ability to recursively self-improve and adapt poses an existential threat to humanity – and that Anthropic’s employees know it.
“Do not underestimate the power of this technology. These will soon be superhuman systems that can hack anything, revolutionize any field overnight, and acquire real power and resources,” Coxon wrote. “The people building AI earnestly believe that it could kill us all by the end of the decade… No other human activity poses this level of danger.”
The warning took the tech world by storm, receiving significant social and mainstream media attention. A current Anthropic employee named Evan Hubinger responded to Coxon’s post, validating his concerns with strong language.
“Jacob is correct here—we really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade,” Hubinger replied on X. “I believe Anthropic is trying its best, but we do not yet have a plan to solve alignment for superintelligence and are not clearly on track to.”
These comments echo the very revolt that led to Anthropic in the first place. The company was founded by former OpenAI employees who raised red flags about security, but deployments of its own products have often been coupled with warnings that the technology is developing too fast. In late July, Anthropic announced that it was investigating three incidents in which AI agents it was developing were able to autonomously access the open internet and hack external parties. Yesterday, the company reported a fourth such incident.
Hours prior to this article’s publication, Anthropic released another report stating that the company intervened on research done by scientists using their platform that could have led to the development of biological weapons.
Cambridge officials react
While Coxon and Hubinger’s claims have sparked debates among Washington lawmakers over how to regulate the technology, some city officials highlighted the economic upside to having Anthropic in Cambridge long-term.
“The City of Cambridge is very excited about this announcement and welcomes the expansion of Anthropic into Kendall Square,” said Director of Media Relations Jeremy Warnick. “Extending their presence in the community reinforces Cambridge as one of the best ecosystems for the AI and Quantum industries.”
The lease could be a positive sign for commercial property tax revenues, which municipalities across the Commonwealth have fretted over as commercial office and lab space vacancies risk depressing property values. City Manager Yi-An Huang told Cambridge Day earlier this year that the city was getting close to hitting its levy limit, the total percent it is legally allowed to tax under state law, on commercial properties.
News of the forthcoming announcement was not widely known across city staff, but Warnick says employees from the Economic Opportunity and Development office had heard from state and real estate executives that Anthropic was close to signing a lease.
City councillors Patty Nolan and Jivan Sobrinho-Wheeler both said they learned of the development from headlines. Sobrinho-Wheeler countered claims that an increased Anthropic presence would be an economic boon for the city.
“The news of a big AI company that says it’s going to put people out of work setting up shop in Cambridge is one more reason we need the Cambridge Jobs Training Trust I’ve been pushing for,” said Sobrinho-Wheeler, referencing comments made by Anthropic CEO Dario Amodei on the labor impacts of AI.
Sombrinho-Wheeler has advocated for a model similar to ones implemented in Boston and Somerville in which developers of large commercial properties pay a fee that goes towards jobs training programs.
“We need to be looking out for our residents’ livelihoods because corporations like [Anthropic] are not.”
A stronger
Please consider making a financial contribution to maintain, expand and improve Cambridge Day.
We are now a 501(c)(3) nonprofit and all donations are tax deductible.
Please consider a recurring contribution.