#QuitGPT: the mass exodus from ChatGPT to Claude has begun

On February 28, 2026, OpenAI signed a deal with the U.S. Department of Defense.

Within hours, the hashtag #QuitGPT exploded across social networks worldwide.

In one week, 2.5 million users pledged to cancel their ChatGPT subscription or delete the app.

It’s the largest organized boycott in AI history, and it has one very specific destination: Anthropic’s Claude.

A military deal that set everything in motion

The backstory starts several weeks before the movement erupted.

On February 27, 2026, the U.S. Department of Defense publicly revealed it had approached both OpenAI and Anthropic about providing AI services for military systems.

The two companies gave diametrically opposite responses.

OpenAI says yes, Anthropic says no: two irreconcilable visions

Dario Amodei, CEO of Anthropic, published an unambiguous public statement:

I cannot in good conscience comply with the Pentagon’s request regarding domestic mass surveillance or fully autonomous weapons.

Anthropic turned down a contract estimated at $200 million, arguing that current AI models are not reliable enough for autonomous weapons systems and that mass surveillance of American citizens constitutes a violation of fundamental rights.

OpenAI signed the deal the following day.

Sam Altman stated the contract includes the same limitations Anthropic had demanded, but the legal wording differs on several critical points.

The Pentagon went further, designating Anthropic as a “supply chain risk” and ordering all its contractors to stop using Anthropic models for military projects.

Anthropic has since filed lawsuits against the Department of Defense and several federal agencies, challenging this designation imposed by the Trump administration.

The public response: 2.5 million people vote with their feet

Users didn’t wait long to react.

Daily uninstalls of the ChatGPT app spiked 295% above their usual average.

The #QuitGPT movement crystallized several grievances that had been building for months:

• The Pentagon deal covering AI military systems and surveillance
• OpenAI president Greg Brockman’s $25 million in donations to MAGA Inc., the pro-Trump super PAC
• The use of ChatGPT-4 by ICE (Immigration and Customs Enforcement) to automatically screen résumés as part of deportation proceedings

ChatGPT’s paying user base (developers, researchers, writers, knowledge workers) is precisely the audience most sensitive to questions of AI ethics and corporate accountability.

It’s also the audience that knows Claude, knows how to make the switch, and has the technical skills to migrate without friction.

Why Claude became the natural destination

The movement could have splintered toward dozens of alternatives: Gemini, Mistral, open-source solutions.

Instead, it converged heavily on Claude, and that’s no accident.

When Claude topped ChatGPT in the U.S. App Store for the first time, Anthropic immediately launched a migration tool to import your complete ChatGPT conversation history in just minutes.

Ethics by design, not just a marketing angle

Anthropic built its model around an internal constitution of roughly 30,000 words that guides Claude’s behavior across thousands of edge cases.

On privacy, the difference with ChatGPT is structural:

• Claude: your conversations are not reused for training by default, with no premium subscription required
• Standard ChatGPT: conversations can be reviewed by OpenAI and used for training, unless you manually disable it in settings
• Anthropic doesn’t sell any data to advertisers and doesn’t monetize user activity

For users looking to make the switch, our detailed guide on migrating your data from ChatGPT to Claude walks through every step of the process.

The performance that won over even the skeptics

Ethics alone wouldn’t be enough if Claude were technically inferior.

On the SWE-bench Verified benchmark (resolving real bugs in open-source repositories), Claude Opus 4.6 scores 80.8% against roughly 80% for GPT-5.3 Codex.

Performance is now equivalent or slightly better for coding use cases, which account for a large share of the paying user base at $20 per month.

Our complete comparison helps you choose the best AI model for your projects based on your specific needs.

What the migration to Claude tells us about the future of AI

This movement exposes a reality the AI industry has long ignored: users are not passive.

When a tech company crosses a perceived ethical line, a portion of its base can shift to a competitor, and shift fast.

#QuitGPT is not a fringe ideological movement: it’s a market signal that AI companies will need to learn how to read.

The numbers behind a historic shift

ChatGPT‘s share of the AI assistant market dropped from 60% in early 2025 to under 45% in Q1 2026.

Claude went from 8% to 18% over the same period.

On Anthropic’s side, internal figures confirm the trend:

• Daily sign-ups hit all-time highs following the Pentagon refusal announcement
• Free users surged: +60% since January 2026
• Paid subscribers more than doubled over the full year

Claude overtook ChatGPT in the U.S. App Store rankings the very day after Anthropic’s refusal: an absolute first since both apps launched.

Will this movement last

Tech boycotts often follow a familiar pattern: an initial surge, then a partial return to old habits.

#QuitGPT has several factors that set it apart from previous attempts.

The landmark tech migrations of the past (Yahoo to Google, MySpace to Facebook) cemented themselves because the alternative was objectively better across every dimension, not just one.

Here, Claude isn’t just the ethical choice: it’s an option that competes on performance, offers comparable pricing, and provides ready-made migration tools.

The structural decline in ChatGPT’s market share started before the boycott: #QuitGPT simply accelerated a shift already underway.

Even as users exit, OpenAI continues to expand into new verticals: its ChatGPT Health initiative shows the company is betting on specialized AI assistants rather than reversing its strategic direction.

For the technical context behind this clash, our analysis of the latest GPT-5 updates puts each model in perspective.

Conclusion

#QuitGPT marks a turning point in the history of mainstream AI.

For the first time, millions of users left the market leader: not because a competitor scored higher on a benchmark, but because one company publicly chose an ethical position and held it under considerable financial and governmental pressure.

Anthropic turned down $200 million from the Pentagon, and that decision earned them something money can’t buy directly: the trust of a demanding user community.

Every tech company should pay attention to this signal: in a sector where products are converging technically, values become a real, measurable competitive advantage.

If you’re considering switching to Claude, now is a good time: the migration tools exist, the performance is there, and the community has never been larger.

FAQ

What is the #QuitGPT movement

#QuitGPT is a boycott movement launched in late February 2026 after OpenAI signed a deal with the U.S. Department of Defense.

Within one week, 2.5 million users pledged to cancel their ChatGPT subscription or delete the app, making it the largest organized boycott in AI history.

Why did Anthropic refuse the Pentagon contract

Anthropic turned down a contract estimated at $200 million because the Pentagon wanted to use AI for mass surveillance of American citizens and fully autonomous weapons systems.

CEO Dario Amodei deemed these use cases incompatible with Anthropic’s values and with the current reliability level of AI models.

Is Claude better than ChatGPT

On the SWE-bench Verified benchmark, Claude Opus 4.6 scores 80.8% against roughly 80% for GPT-5.3 Codex.

Both models perform similarly overall, but Claude stands out on privacy by default and transparency around data usage.

How do you migrate from ChatGPT to Claude

Anthropic launched an official migration tool that imports conversation history and preferences from ChatGPT in just a few minutes.

The tool is accessible directly from the Claude.ai interface.

Does Claude protect privacy better than ChatGPT

Yes: Claude does not reuse your conversations to train its models by default, and no premium subscription is required to get this protection.

The standard version of ChatGPT allows OpenAI to review and use conversations for training unless you manually disable this in settings.

What is the “supply chain risk” designation imposed on Anthropic

After Anthropic’s refusal, the Pentagon designated the company as a “supply chain risk”, barring all its contractors from using Anthropic models in military projects.

Anthropic challenged the designation by filing lawsuits against the Department of Defense on March 9, 2026.

Are there other reasons behind the boycott beyond the military contract

Yes: FEC filings reveal that OpenAI president Greg Brockman and his wife each donated $12.5 million to MAGA Inc., the pro-Trump super PAC.

The use of ChatGPT-4 by ICE to automatically screen résumés as part of immigration deportation proceedings also fueled the movement.

How many users did Claude gain from #QuitGPT

According to Anthropic, daily sign-ups hit all-time highs, free users grew by 60% since January 2026, and paid subscribers more than doubled.

Claude surpassed ChatGPT in the U.S. App Store rankings for the first time since both apps launched.

What is Claude’s current market share compared to ChatGPT

ChatGPT‘s share went from 60% to under 45% between early 2025 and Q1 2026.

Claude went from 8% to 18% over the same period, a progression that started before and extends beyond the boycott alone.

Will the #QuitGPT movement last

Experts note that #QuitGPT stands apart from previous tech boycotts because the alternative is competitive on performance, offers a straightforward migration path, and benefits from a consistent ethical positioning.

The structural decline in ChatGPT’s market share started before the boycott: #QuitGPT accelerated a migration already in progress.