What Happened
Mark Hurst, founder of the long-running Creative Good newsletter (established 1998), published a sharp editorial on March 5, 2026 titled "AI and the Illegal War." The piece directly connects consumer AI companies to military applications and argues that the industry's pursuit of government defense contracts undermines its public-facing mission of beneficial AI.
Hurst's central claims: AI systems are being integrated into military targeting workflows through programs like the Maven Smart System, with major AI companies providing the underlying technology. He specifically names Anthropic's Claude as being involved in target identification and coordinate generation for military operations.
The piece also cites Goldman Sachs data indicating that massive AI investment "contributed basically zero" to US economic growth in 2025, arguing that military contracts represent a revenue fallback for companies whose commercial AI products have not generated the returns investors expected.
Hurst characterizes the current dynamic as one where tech billionaires are leveraging AI for military revenue while marketing the same underlying technology as productivity tools to consumers.
Why It Matters
If you use AI tools from any major provider, the question of where else that technology gets deployed is not abstract. Military contracts can influence product development priorities, data handling practices, and company culture in ways that eventually affect consumer products.
The tension between commercial AI and defense AI is real. Companies like OpenAI, Google, and Anthropic have all faced internal debates about military work. OpenAI quietly removed its ban on military applications in early 2024. Google's Project Maven controversy dates back to 2018. These decisions shape the organizations building the tools millions rely on.
For daily AI tool users, the practical concern is less about the ethics of any single contract and more about the trajectory. If defense revenue becomes a primary growth driver for AI companies, product roadmaps may shift toward capabilities that serve government clients rather than individual productivity users.
Our Take
Hurst's piece is an opinion editorial from a writer with a clear editorial stance, and should be read as such. The specific claims about military operations deserve scrutiny and verification beyond a single newsletter source.
That said, the broader point about AI companies pursuing military contracts is well-documented and worth paying attention to. The transition from "we will never do military work" to "we will do defensive military work" to active defense contracting has happened at multiple AI companies over the past two years. That pattern is real regardless of one's view on whether it is justified.
For AI tool users, the honest assessment is this: the companies building your productivity tools are also building military tools. Whether that bothers you is a personal judgment. But you should at least be aware of it when evaluating which companies to invest your workflows in.
The Goldman Sachs data point about AI's limited economic contribution is perhaps more relevant to this audience. If the commercial AI market has not delivered the growth investors expected, the pivot toward government and military contracts could accelerate. That would mean the tools we use daily are subsidized by defense spending - a dynamic that has shaped the tech industry before (the internet itself started as a military project) but one worth understanding clearly.