Generative AI company Anthropic's tighter usage limits for Claude are changing how some users work. The practice is spreading among founders and developers on lower-priced subscription plans of breaking work into smaller parts and planning schedules around limit reset times.

Business Insider reported on April 13 that Anthropic adjusted Claude's usage limits in late March to manage peak-time demand. The company said it caps usage based on five-hour sessions and is working on efficiency measures to meet surging demand. It said the change could affect about 7 percent of users who previously did not hit limits.

The change is directly affecting users who rely on AI as a work tool. Max Johnson (맥스 존슨), a co-founder of British startup Brix, said he used to handle content writing, design and document work in a single long Claude chat. Recently, he has found himself reaching the limit after entering only a few prompts even after opening a new chat, he said. He said he now plans his day with an "invisible meter" in mind.

Work patterns have also shifted. Instead of keeping a long context, it has become common to split tasks into social media scripts, graphics and documents and handle each in a separate chat. Users are also narrowing the scope of instructions to reduce token usage. Some teams have switched from shared accounts to individual accounts and are also considering adopting an enterprise plan.

Session limits can create work gaps beyond a simple inconvenience. Johnson said work stops if team members hit the limit at the same time, forcing them to redraw plans for the next steps. "That is when panic hits," he said, adding that he sometimes spends about 30 minutes to 1 hour a day thinking through the next steps.

Users are adapting to a new work rhythm by rebuilding schedules around limit reset times. Some developers manage AI usage like a "weekly budget", concentrating high-load tasks when they have room under the cap and lowering work intensity as they near the limit, it said.

Ani Potts (애니 포츠), a New York University mathematics student preparing to launch a startup, said she schedules demanding tasks such as research, testing and coding when she has room under the AI usage cap. When she gets close to the limit, she stops or shifts to less demanding issues. Some say the time spent not using AI instead helps organise their thinking.

Others say limits could lead to lower productivity. A developer in Canada said hitting the cap often means stopping work on a project. Some also said concentrating AI use into short periods can reduce fatigue.

Ultimately, Claude's usage caps are being seen as reshaping not only subscription terms but also how users who rely on AI as a daily productivity tool allocate time and structure work. The impact is greater for regular subscribers than for enterprise plan users, drawing attention to how AI service pricing policies and usage structures could affect the overall productivity environment.