Anthropic Flags Data Center Backlash as Corporate Risk

Anthropic says opposition to data center development is becoming more organized and coordinated, and warns that its public profile may draw extra attention to projects associated with the company. The warning appears in its IPO prospectus, alongside risks tied to energy demand, competition, child safety and the possibility that customers will reject the price of frontier AI models.

Organized Opposition Meets Corporate Expansion

“Opposition to data center development has also become more organized and coordinated across jurisdictions and our public profile may draw particular attention to projects associated with us,” the filing said. That’s the prospectus’s account of people organizing around construction that the company depends on to train and maintain its models. It offers no account of who those people are or what they want; it does make clear that their opposition has become a risk for the business.

The prospectus, revealed Monday, runs for more than 90 pages. It says lawmakers in 16 states are considering moratoriums on data center construction as midterm elections approach and construction becomes a political issue. The filing presents those proposals and public opposition as threats to Anthropic’s plans. It also shows the tension plainly: an expanding portfolio of computing facilities is central to the company’s operations, while organized resistance can put projects under scrutiny.

Anthropic’s filing does not describe a community response in detail, name organizers, or quote anyone opposing construction. There’s no account of mutual aid or horizontal organizing, either. What it does record is a corporate concern that organized opposition across jurisdictions could affect projects tied to the company.

The Infrastructure Bill Comes Due

Anthropic says its prospects depend on electrical and related power infrastructure. A lack of “sufficient electrical power” across the nation and the world poses a risk to the company and its rivals. The filing also warns that power could become more expensive even if supply grows through “on-site natural gas-fired generation,” because demand is high.

That’s the infrastructure calculation in the prospectus: more computing facilities require power, and higher demand could raise its cost. The document identifies the risks to the company’s business, but provides no figures for energy use or prices and no account of how costs might affect people outside the company. Its focus is the corporate balance sheet and the supply needed to keep models running.

Anthropic also warns that demand for its models, platform and applications could develop more slowly than expected or decline, slowing or stalling growth. Customers may turn to cheaper or open-source models if they decide “incremental performance gains at the frontier do not justify the additional cost.” The company’s market depends on businesses continuing to spend on AI. That spending isn’t guaranteed.

Safety Risks, Written in Corporate Terms

The prospectus says Anthropic’s models could cause “catastrophic or existential risks to humanity.” It also says the complexity of the technology prevents the company from guaranteeing that its safety efforts will be adequate or predicting every potential scenario. Fierce competition, market pressure and stakeholder expectations, Anthropic says, make safe development and implementation challenging. It warns that failing to emphasize AI safety could hurt its “brand, business, financial condition, and results of operations.”

The filing also points to legal risks involving children. OpenAI and Meta have faced scrutiny from lawmakers and plaintiffs’ attorneys over allegations that the design and implementation of their AI models and applications contributed to child-safety-related harms. Meta recently agreed to pay roughly $17 billion in a settlement with multiple state attorneys general in litigation involving child mental health and technology addiction.

Anthropic says it has largely avoided scrutiny so far but could face legal challenges. Its terms of service bar minors from using Claude and related products. The company says it must protect children “as adoption of our models grows, including by minors and other vulnerable users, notwithstanding our efforts to restrict such access.” The prospectus frames that duty as a risk to the business, too.