Anthropic Doubles Claude Usage Limits for All Users Through March 27
Anthropic launched a two-week promotion giving every Claude user — Free, Pro ($20/mo), Max ($100–$200/mo), and Team ($30/user/mo) — double their normal usage limits during off-peak hours. The 2x boost activates automatically outside weekday peak hours (8am–2pm ET) and runs all day on weekends. No code, no toggle — it just happens.
The move serves two purposes. First, it smooths out Anthropic’s compute load by incentivizing off-peak usage. Second, it gives power users who’ve been hitting rate limits a reason to stay on Claude instead of switching to competitors during heavy sessions. For developers on the API, this doesn’t change API rate limits — it’s a consumer/team product promotion only. The promotion runs March 13–27.
Claude Code Gets 1M Context on All Plans, New UI Features
The latest Claude Code update brings the full 1M token context window to Opus 4.6 across all plans — previously this required Max or Enterprise. The update also adds a /color command for customizing the terminal interface, displays the session name on the prompt bar for easier multi-session management, and now shows memory freshness timestamps so you can see when Claude last updated its project context.
These are quality-of-life improvements rather than major feature drops, but they address real friction points. The 1M context availability is the big one — it means Claude Code users on Pro can now work with entire codebases in a single session without running into the old 200K ceiling.
OpenAI and DeepMind Researchers File Amicus Brief Supporting Anthropic
Researchers from OpenAI and Google DeepMind have filed an amicus brief in personal capacities supporting Anthropic’s lawsuit against the Pentagon. The brief argues that the supply chain risk designation sets a dangerous precedent for the entire AI industry — if the government can blacklist a company for setting ethical boundaries on military use, it creates a chilling effect on responsible AI development across the board.
This is notable because OpenAI and Anthropic are direct competitors. The fact that individual researchers from rival companies are willing to publicly support Anthropic’s legal position shows how seriously the AI research community views the Pentagon’s overreach. Legal experts cited by StartupNews say Anthropic has a strong case, particularly on First Amendment grounds.
US Army Awards Anduril $20 Billion Defense Tech Contract
The US Army announced a 10-year contract with Anduril Industries worth up to $20 billion, covering autonomous systems, counter-drone technology, and AI-powered command and control. The deal is one of the largest defense tech contracts ever awarded to a non-traditional contractor and positions Anduril as a primary Pentagon AI partner going forward.
The timing matters. This contract was announced the same week Palmer Luckey publicly backed the Pentagon’s Anthropic blacklist. Anduril is the clearest beneficiary of Anthropic’s removal from the defense ecosystem — every AI workflow that loses Claude will need a replacement, and Anduril is already embedded across multiple military branches. Whether the contract and the blacklist are directly related or not, the optics tell a story.
Palmer Luckey: Pentagon Should Have Been “More Forceful” With Anthropic
Anduril founder Palmer Luckey told Axios in an exclusive interview that the Pentagon should have been even more aggressive in its response to Anthropic’s refusal to support military AI applications. Luckey argued that private companies shouldn’t get to dictate how the US military uses their technology once it’s been contracted, and that Anthropic’s ethical stance on military AI amounted to undermining national security.
Luckey’s comments are the sharpest public criticism of Anthropic from a defense industry leader to date. They also reveal the philosophical divide in the AI industry: Anthropic believes companies have a responsibility to set ethical boundaries on their technology’s use; Luckey and the defense establishment believe that once you take government money, the government decides how your tools get deployed. With Anduril freshly awarded a $20B Army contract and Anthropic fighting a blacklist in court, these two visions of AI governance are heading for a collision.
The Off-Peak Play: Anthropic Is Solving Its Compute Problem in Public
Doubling usage limits during off-peak hours looks like a generous promotion, but it’s really an infrastructure optimization move. Anthropic’s compute capacity is heavily loaded during US business hours — the rate limits exist because demand genuinely exceeds supply during peak. By offering 2x limits off-peak, they’re incentivizing users to shift their heavy usage to evenings, early mornings, and weekends when GPUs are underutilized.
This is the same playbook utilities have used for decades with time-of-use pricing. The interesting question is whether it becomes permanent. If enough users shift their workflows to off-peak — especially the Max subscribers running long coding sessions — Anthropic gets better utilization from existing hardware without buying more compute. For users, it’s a genuine win: more Claude for the same price, with the only cost being a schedule adjustment. Watch for this to become a standard feature rather than a two-week promotion.