Anthropic is doubling Claude's five-hour usage limits during off-peak hours from March 13 through March 27, 2026, in a limited-time promotion that extends across web, desktop, mobile, and enterprise integrations including Claude Code and Excel. Enterprise plan subscribers are excluded, though the doubled limits apply to Free, Pro, Max, and Team plans across the platform.
On its surface, the offer reads as a thank-you to Claude's rapidly expanding user base. The promotion was announced as a token of appreciation for Anthropic's growing user base, and the doubled usage is automatically applied during off-peak hours with no action required from users. But the timing reveals deeper competitive pressures shaping the AI market.
Claude's momentum has surged in recent weeks. In the short term, the clash with the Pentagon has fuelled interest in Claude, as some social media users call for dumping ChatGPT over OpenAI's deal with the Pentagon. This popular push has translated to real adoption; Claude reached the top of app store charts in late February despite the company's broader legal and regulatory troubles.
The Pentagon conflict looms large beneath this promotion, however. The Trump administration ordered federal agencies and contractors that work with the military to cease business with Anthropic after the company refused to allow the Pentagon to use its artificial intelligence technology without restrictions, with a six-month phase-out period. Within Anthropic's contract with the Pentagon were prohibitions against the use of Claude in mass surveillance and autonomous weapons.
The real danger to Anthropic's business comes not from lost government revenue, but from enterprise churn. Much of Anthropic's success stems from its enterprise contracts with big companies, many of which may have contracts with the Pentagon, meaning that Anthropic's existing customer base might evaporate, either because they have government contracts or might want them in the future. Early signs suggest this risk is already materialising. Ten venture-backed portfolio companies working with the Department of Defense have backed off their use of Claude for defence use cases and are in active processes to replace the service with another one.
The off-peak promotion tackles two problems at once. First, it gives free users genuine access to Claude's full capabilities during the two-week window. Free-tier users typically run into rate limits fast, sometimes within a handful of longer conversations, but this promotion loosens those constraints enough that someone could realistically test Claude on actual work tasks rather than toy examples. For developers and knowledge workers evaluating AI tools, this effectively removes a major friction point.
Second, the move signals that Anthropic has infrastructure headroom during non-peak hours. Anthropic has enough spare capacity during off-peak windows to give it away, which suggests either strong infrastructure buildout or lower-than-expected off-peak demand. In a market where compute capacity is scarce and expensive, the ability to sacrifice revenue on off-peak usage suggests confidence in overall infrastructure investment.
The broader context matters for Australian readers considering Claude adoption. Losing a $200 million contract would not pose an existential threat for Anthropic, which was recently valued at around $380 billion, but the bigger risk is the supply chain risk designation, which means any company that works with the US military would have to prove they don't touch anything related to Anthropic in their work with the Pentagon. For Australian enterprises with US defence connections, this creates uncertainty around Claude's long-term viability in their technology stack.
The AI assistant market is brutal right now. OpenAI's ChatGPT still dominates mindshare, Google's Gemini is baked into everything from Gmail to Android, and Microsoft's Copilot ships with Windows, while Anthropic, despite raising over $7.3 billion in funding, doesn't have a default distribution channel. The usage promotion is a calculated bet that if enough new users try Claude's capabilities during these two weeks, a portion will convert to paying customers before the limitation resets on March 28.