With all the discussion around artificial intelligence, some still view AI as abstract, but make no mistake: Artificial intelligence is the new, and very real, engine that will power our economic and military might in the future.
We’ve built strong AI models on foundational AI chip technology, but our success in the AI race against China will be decided less by the models themselves and more by what we produce through them.
Recommended Stories
Think of AI models as the golden goose and the applications as the real golden eggs.
Companies such as Anthropic, which own closed-source models such as Claude, know that coders will be captive to them as long as they’re the biggest game in town.
Claude’s closed-source development code is wrapped in a black box that not even Anthropic CEO Dario Amodei’s mother could open — a box that’s only accessible through a metered payment plan based on tokens instead of an all-inclusive monthly fee. And the third-party tools, such as OpenClaw, that could have helped Claude make apps more cheaply and effectively? Anthropic blocked them. All of this in the name of “safety.”
Anthropic’s restrictions are suffocating innovation just as America is seeing an AI app explosion fueled by entrepreneurs with little to no programming background.
Thousands of U.S. companies are trying to build healthcare, logistics, defense, and finance apps. Stifling advancement with unnecessarily restrictive code, as Anthropic has done, freezes our momentum at exactly the wrong moment, dragging America down in the global AI race.
In the meantime, Chinese tech firms are making their own AI platforms open source to lighten the burden on innovation. They’re opening major elements of their black box so experts can tinker with and improve them. App developers using China’s AI products and services get much more control and flexibility over their final products using open-source models — plus a massive cost savings advantage because they’re much cheaper to use.
Letting loose open-source models is a genius move by China in a larger chess match over AI dominance. Making High-Flyer’s Deepseek, Alibaba’s Qwen series, and Baidu’s Ernie open-source helps a new generation of startups turn advanced AI into a shared foundation of apps that we can all build on. Anthropic risks sacrificing more of America’s chess pieces the longer it waits to respond.
Open-source code is sprinkled across 96% of all commercial software. This is the same kind of decentralization, competition, and iteration that America has thrived on — and that Chinese tech firms have now adopted and used against us.
Open-source code has saved commercial software developers nearly $9 trillion. Companies such as Anthropic know this, which is why they’re trying to keep customers hooked on Claude’s ecosystem. They selfishly want a larger piece of those trillions by making innovators pay through the nose up front in tokens.
Meanwhile, Anthropic and other big players are trying to slow down new competitors by reshaping the environment under the guise of national security. Amodei and Google DeepMind’s Demis Hassabis are pushing for a U.S.-led standards body that would test AI models for national security risks. It’s a classic big-company move to protect its position. They’re trying to add one more lockable door that smaller competitors with no key must pass through.
Anthropic’s security arguments are a facade of concern padded with profit motives. America has a strong record of managing risks, and we’ve been building open-source computing safety nets for decades.
Anthropic itself benefits from open-source code. Claude uses Python extensively when running complicated tasks. It even donated $1.5 million to the Python Software Foundation to make Python more secure. They clearly see the value of open-source code. Now they should fully embrace it before Chinese firms zip past us in the AI race.
Yes, we need software safeguards, but they should be targeted. Anthropic’s have-my-cake-and-eat-it-too approach to open source will likely lead to real consequences for America. It will mean fewer startups, slower innovation, and tech talent leaving the United States for, ironically, more innovative frontiers.
RED TAPE WILL KILL THE NEXT AI STARTUP, AND BIG TECH COULDN’T BE HAPPIER
This new wave of application development has stakes for all of us. We want better healthcare and innovative new drugs. We also want less stressful workdays. And if we don’t get good products or services at a fair price, we want the freedom to choose alternatives.
Policymakers should think twice before letting companies such as Anthropic write the rules for their own industry. They’re just trying to push an ideological agenda with selfish motives — all while killing growth and innovation in America to China’s benefit.
Stephen Moore is an economist and co-founder of the Committee to Unleash Prosperity.
