The AI Arms Race and Emerging US Parallels to China
American AI companies including OpenAI, Anthropic, Google, Microsoft, and Meta frequently cite the need to outpace China in AI development to avoid ceding global superpower status to an authoritarian regime. China's government employs AI for suppression and surveillance, a model the US rightly opposes. However, recent Pentagon actions against Anthropic suggest the US may be adopting similar coercive strategies domestically.
The Department of Defense had a $200 million contract with Anthropic for its Claude AI, used in operations like the January raid in Venezuela. Disputes arose over Anthropic's firm red lines: no mass domestic surveillance or fully autonomous weapons. The Pentagon demanded unrestricted access, threatening to designate Anthropic a supply chain risk—a label previously reserved for foreign adversaries like Huawei.
The Pentagon’s threats against Anthropic copy the worst aspects of China’s military-civil fusion strategy. China’s actions to force high-tech private companies into military obligations may lead to short-term technology transfer, but it undermines the trust necessary for long-term partnerships between the commercial and defense sectors.
OpenAI Steps In with Compromised Terms
Hours after blacklisting Anthropic, OpenAI announced deployment of its models on the Pentagon's classified network. CEO Sam Altman claims alignment with Anthropic's red lines, yet the deal permits use for 'all lawful purposes.' Current US laws lag behind AI capabilities, allowing government purchase of private data for mass analysis—geolocation, browsing, credit card info—creating predictive profiles without explicit surveillance bans.
Anthropic viewed such data practices as unacceptable; OpenAI's contract excerpt confirms broad lawful use. Additional clauses offer illusory protections, as experts note legalese enables circumvention. Technical guardrails and embedded engineers provide limited assurance, especially for complex military applications.
In terms of safety guardrails for ‘high-stake decisions’ or surveillance, the existing guardrails for generative AI are deeply lacking, and it has been shown how easily compromised they are, intentionally or inadvertently.
Nothing in the contractual language released up to this point seems to provide enforceable red lines beyond having a ‘lawful purpose.’ Embedding OpenAI engineers does not solve the problem.
Public Backlash and Boycott Momentum
OpenAI faces swift public rejection via the QuitGPT campaign, claiming over 1.5 million participants urging switches to alternatives like Anthropic's Claude, now the top App Store download. Historian Rutger Bregman praises its targeted focus. Even OpenAI's Leo Gao questioned internal commitments publicly.
Anthropic is not flawless—partnering with Palantir for ICE operations and relaxing prior safety promises—but stands firmer against Pentagon overreach.
What effective boycotts have in common, in my view, is that they’re narrow, they’re targeted, and they’re easy. This is the first opportunity to start a massive consumer boycott in the AI era, and to send an incredibly powerful signal to the whole ecosystem, saying, ‘Behave, or you could be next.’
Paths Forward: Solidarity and Regulation
Tech workers respond with open letters: over 900 signatures from OpenAI and Google employees pledging solidarity against surveillance and autonomous weapons; another 175 from industry leaders urging Pentagon reversal and Congressional review. International treaties remain elusive despite Nobel-backed calls. Cross-company coordination offers interim resistance to pressures mimicking the systems the US condemns.
