You paste a block of code into a chat window and, for a moment, your history flickers back to life. I watched Claude climb the charts while debates over military deals turned into a public brawl. You can feel the stakes: your data, your preferences, and potentially the future of where these models live.
On Saturday, Claude reached the top of Apple’s App Store — and downloads didn’t stop
I want you to imagine the moment an app moves from curiosity to habit. For Anthropic’s Claude, a single weekend of headlines and a fresh import tool pushed it past OpenAI’s ChatGPT in the free apps chart.
The new feature promises a painless handoff: paid users can ask another chatbot to export a chunk of memory in a code block, paste it into Claude’s Settings → Capabilities → Memory → “Start import,” and press “Add to memory.” The result is immediate continuity: preferences, context, and prior prompts follow you into Claude like swapping out the engine in mid-race.
How do I import memories from ChatGPT to Claude?
You copy Anthropic’s export prompt into the other model, request the memory export in a code block, then paste that block into Claude’s memory importer. It’s intentionally manual—one paste—and it avoids rebuilding long histories by hand. Google is testing similar moves for Gemini, and ChatGPT already offers ways to transfer conversations between ChatGPT accounts.
On Friday, talks with the Defense Department broke down and a political fight spilled online
I’m telling you this because the controversy around military use is not background noise — it changed the market. Anthropic pushed for exceptions banning use of its models in mass domestic surveillance and fully autonomous weapons. The Defense Department pushed back and, after negotiations failed, Secretary Pete Hegseth said the department would begin treating Anthropic as a supply-chain risk.
That designation would bar contractors and suppliers working with the U.S. military from maintaining commercial ties with Anthropic, a move that immediately rippled through enterprise planning. Anthropic says it made the demand because it believes current frontier models are not ready for those applications and that mass surveillance violates basic rights.
Why did Claude overtake ChatGPT on the App Store?
Public controversy plus a tool that removes friction for users equals momentum. Downloads spiked after the Anthropic-Defense Department standoff, and a boycott movement called “QuitGPT” claimed more than 1.5 million participants. Celebrity support — Katy Perry posted a screenshot endorsing Claude Pro — and partisan reactions, including a Truth Social post from Donald Trump, kept the story trending.
On Monday morning, Claude experienced outages as demand surged
I watched status pages and user reports; Anthropic blamed “unprecedented demand.” OpenAI moved fast: within hours of the blacklisting, OpenAI announced an agreement with the Defense Department to deploy tech in classified systems. Elon Musk’s xAI reportedly agreed to similar terms, and systems from Grok, Google, and OpenAI appear on GenAI.mil for unclassified uses.
OpenAI also claims its contract protects the same prohibitions Anthropic demanded, including bans on mass surveillance and fully autonomous weapons — a claim that will get pulled apart in hearings, contracts, and press cycles. Meanwhile, users debating ethics and convenience are deciding which ecosystem to trust.
On hundreds of threads, people asked whether this matters for daily use
You don’t need a government contract to care about portability. If you’ve built years of preferences into ChatGPT or another LLM, the idea of starting over is painful. Anthropic’s importer is a small but powerful gesture toward user control — a digital passport that carries your habits and context from one model to another.
For paid Claude users, the path is: generate the export in the other chatbot, copy the code block, paste into Claude’s importer, and confirm. For people considering whether to join QuitGPT, headlines about Greg Brockman’s reported $25,000,000 (€23,000,000) donation to a pro-Trump super PAC added fuel.
On the industry stage, the real contest is between policy, product, and loyalty
I’m watching three threads: product features that reduce switching costs, political and defense deals that shape vendor access, and community reactions that drive consumer behavior. Anthropic has leaned into consumer goodwill with an explicit import tool while OpenAI and xAI maneuver in government lanes. Google’s Gemini is experimenting in the same space, and ChatGPT keeps refining account transfers.
Which firms win will depend on trust and convenience — not just raw capability. You care about control and continuity; companies care about contracts and scale. That tension is now public and messy, and it will rewire which models businesses and everyday users choose.
So where do you place your bets — convenience, conscience, or contract?