Anthropic: The scrappy underdog punching above its weight and willing to fight
The 2026 AI & election tech forecast looks at how the AI company is trying to differentiate itself in DC
This post is part of “The AI x Election Tech Forecast,” a limited series unpacking how AI is reshaping the role and responsibility of major tech platforms in the 2026 election cycle. Each edition highlights the shifts happening now before the rest of us catch up. See the whole series here.
Editor’s note: This analysis was developed through collaboration with Claude from Anthropic. My process: I provided a brain dump of my strategic insights and insider knowledge. I then went back and forth with the tool on the structure and framework. Claude asked clarifying questions to fill gaps, then drafted the piece based on my answers. I edited and fact-checked the final output. For more on how I use AI see my ethics and transparency disclosure.
📖 1375 words · ~5 min read
1 big thing: Anthropic zigs when everyone else zags 🥊
While Meta spends $65M on pro-AI candidates and OpenAI’s Sam Altman cozies up to Trump, Anthropic CEO Dario Amodei is calling out the administration’s China chip policy as “selling nuclear weapons to North Korea.”
While OpenAI announces ads, Anthropic runs Super Bowl commercials promising “Ads are coming to AI. But not to Claude.”
While competitors stay quiet, Anthropic publicly opposed Trump’s proposed 10-year moratorium on state AI regulation and watched the Senate vote it down 99-1.
Why this matters: Anthropic is the contrarian willing to speak uncomfortable truths even when it risks government contracts. In an industry racing to appease the Trump administration, they’re the company betting that principle matters more than proximity to power.
The risk: They could very quickly become the ones with the biggest target on their back—particularly if they keep inviting it by weighing in on political topics.
2. Who they are and what they built 🏗️
Founded: January 2021 by former OpenAI executives Dario and Daniela Amodei, along with other OpenAI defectors who wanted to focus on AI safety.
The differentiation: Constitutional AI—a set of guidelines governing Claude’s behavior focusing on safety, ethics, and being helpful. Updated January 21, 2026, and it’s become their brand.
The growth trajectory: Fastest-growing software company in history. Revenue grew from $1B to $7B run rate over nine months.
The market disruption: When Anthropic released industry-specific Claude add-ons (including legal services), it triggered a dayslong global stock selloff across software, legal services, financial data, and real estate. Then they dropped their most advanced model yet—capable of synthesizing data, running teams of coding assistants, and product management functions. Software company stocks (Salesforce, Intuit) fell again.
Usage: As of late 2025, while ChatGPT has a much larger total user base (900M), Claude is increasingly positioned as superior for enterprise, coding, and complex business automation. 45% of Claude’s API calls come from enterprise users.
3. The Trump tightrope 🎪
Amodei’s position is “cautious collaboration”—agreeing with the administration’s goal of American AI leadership while fundamentally disagreeing on regulation and China policy.
Where they align:
$50 billion American AI infrastructure investment (data centers with Fluidstack in Texas and New York, 800 permanent jobs, 2,400 construction jobs)
Government contracts: $200M Department of Defense agreement, $1 Claude for Enterprise/Gov access across federal government, deployed at classified networks through Palantir and Lawrence Livermore National Laboratory
Where they diverge:
China chip policy: Amodei called Trump’s decision to permit advanced AI chip sales to China “crazy” and a serious national security risk—comparing it to “selling nuclear weapons to North Korea”
Regulation opposition: Opposed Trump’s “One Big Beautiful Bill” 10-year moratorium on state AI laws as too “blunt”—like “ripping out the steering wheel” of AI safety
Only frontier AI company restricting sales to Chinese-controlled companies—forgoing significant revenue to prevent fueling Chinese Communist Party military and intelligence
4. The David Sacks beef 🥩
October 20, 2025: Trump AI and crypto czar David Sacks accused Anthropic of “backdoor Woke AI” regulations through blue states like California, calling them a “Resistance organization” opposing Trump policies.
Sacks’ claim: “It’s likely not a coincidence that the biggest Democrat mega-donors in tech—namely you, Reed Hastings and Dustin Moskovitz—are closely associated with this company.”
October 21, 2025: Amodei responded with a detailed statement saying Anthropic is aligned with the administration on “key areas of AI policy” and interested in working with “anyone serious about getting this right.”
The “woke AI” controversy: White House raised concerns that Claude was biased. Anthropic responded with data:
Manhattan Institute study (conservative think tank): Claude Sonnet 3.5 less politically biased than most other major providers
Stanford study: No reason to single out Anthropic—many other models rated as more biased
System cards for Sonnet 4.5 and Haiku 4.5 show rapid progress toward political neutrality
Status: Ongoing tension, but Amodei continues trying to thread the needle.
5. The Super Bowl shot at OpenAI 📺
February 2026: Anthropic’s Super Bowl ad showed a chatbot offering fitness advice, then slipping in a promotion for shoe inserts for “short kings.”
The punchline: “Ads are coming to AI. But not to Claude.”
Sam Altman’s response: Called it “clearly dishonest” and “on brand for Anthropic doublespeak to use a deceptive ad to critique theoretical deceptive ads that aren’t real.”
Altman’s counterargument: “More Texans use ChatGPT for free than total people use Claude in the US... Anthropic serves an expensive product to rich people. We are glad they do that... but we also feel strongly that we need to bring AI to billions of people who can’t pay for subscriptions.”
My read: This was a short-term branding win that could become a long-term albatross. I’ve seen too many companies make blanket promises like “no ads ever” only to walk them back when market forces or technology evolution demands it. Maybe Anthropic pulls it off. But I’m skeptical.
Anthropic is betting that principle beats proximity to power. Whether that holds is a different question. What matters for 2026 is what they actually did — and what campaigns could do with their tools right now.
Paid subscribers, keep reading for: Anthropic’s 2024 election track record, the campaign tech opportunity most operatives haven’t spotted yet, and my forecast for how long the contrarian bet lasts.
Subscribe to keep reading →
