Anthropic’s People Power Is Part Of A Bigger Fight That Affects Clean Technology

Support CleanTechnica’s work through a Substack subscription or on Stripe.

The Trump administration’s recent decision to label AI startup Anthropic a “supply chain risk” effectively barred the company from federal contracts and isolated it from private firms doing business with the military. While the standoff has been framed largely as a national security issue or a spat with the Trump administration, the resulting fallout and unexpected consumer response highlight critical governance questions for the clean energy and electric vehicle sectors.

The catalyst for the ban stems from a dispute over AI deployment and accountability. Anthropic was integrated into the Pentagon’s systems via a partnership with Palantir. Following a recent military operation in Venezuela, an Anthropic official asked questions about how its AI was used alongside Palantir’s systems during the mission.

Viewing the inquiry as a risk to operational continuity (we have to wonder exactly why, but can only speculate), Palantir escalated the matter to the Pentagon. In response, Secretary of Defense Pete Hegseth issued an ultimatum: Anthropic must drop its safety red lines—which include restrictions on mass domestic surveillance and autonomous weapons—or face removal.

Anthropic refused, prompting the “supply chain risk” designation that cuts it completely out of large swaths of corporate America. Competitors, including OpenAI, subsequently stepped in to fulfill the government’s requirements.

The People Strike Back

The administration’s brash move triggered an unexpected market reaction. As Trump knows well, all news (even bad news) can be good for name recognition and the bottom line. It just didn’t work in his favor this time around.

Anthropic’s consumer app has surged to the #1 spot on app stores, with users subscribing to the company’s $20-a-month Pro tier in record numbers. This citizen-led revenue boost is serving as a financial counterweight to the loss of government contracts, demonstrating a clear market demand for AI models that maintain strict ethical guardrails.

This won’t erase the financial losses that come with being blocked out of the defense world, but it does give the company both a fighting chance and a new constituency that can help it fight in Washington. Anthropic may even get the last laugh when the day comes that both Trump and Hegseth are in the unemployment line in 2027 or 2029.

The Grid Impact of AI Compute

For the cleantech and renewable energy spaces, this precedent regarding software governance and government pressure is highly relevant.

The clean energy transition is increasingly reliant on AI models to manage virtual power plants, balance off-grid solar loads, and optimize charging networks. Yet, the energy required to run these foundational models is staggering.

Training a single large AI model can consume over 1,200 MWh of electricity. Projections show AI could drive global data center energy demand to over 1,000 terawatt-hours by 2026, creating a scenario where aging fossil-fuel plants are kept online simply to meet the tech industry’s compute demands.

If the companies providing the software for grid infrastructure prioritize unchecked expansion and military contracts over efficiency and safety, it fundamentally contradicts the goals of a decarbonized grid. Seeing a company push back, even if they’re not perfect on environmental issues, shows that there’s hope for AI to still be cleaner in the future.

Accountability in Transportation

Beyond the grid, the deployment of AI in transportation has already surfaced serious questions about public safety and corporate accountability.

The automotive industry has witnessed the consequences of prioritizing rapid deployment over rigorous safety testing. Tesla’s “Full Self-Driving” software, deployed as an ongoing public beta, has been linked to numerous crashes, recently resulting in a federal judge upholding a $243 million punitive judgment against the company over an Autopilot fatality. In the robotaxi sector, vehicles have struggled with object permanence, leading to incidents involving pets and pedestrians caught in sensor blind spots.

When AI makes a catastrophic error in transit, regulatory oversight and corporate accountability are often murky. The Anthropic ban sets a precedent wherein a government can successfully pressure a tech company to abandon its internal safety protocols after the product has already been deployed. If the AI models eventually routing autonomous vehicles and balancing public power grids are subject to similar strong-arming or perverse incentives that favor the least ethical players, it introduces a new layer of systemic risk to clean technology infrastructure and electric vehicles.

Final Thoughts

I won’t tell readers to go out and buy an Anthropic subscription, and I won’t tell them not to. It’s a good way to fly Hegseth and Trump the bird, but at the same time, we can’t just ignore the impacts of AI datacenters. We’re all adults and will have to make our own choices on that.

What I will tell you is that Anthropic was right about this particular issue. When government officials wield state power like a little boy who found a .357 in his dad’s sock drawer, handing them even more power in the form of advanced computing technology with no pesky ethical questions asked is definitely a bad call. Worse, it can set terrible precedents that harm basically every other industry that AI touches.