2026-03-24 🧭 Daily News

Pentagon Showdown, iPhone-to-Mac Workflow & $19B Revenue Milestone

🧭 Anthropic vs. Pentagon — The San Francisco Courtroom Showdown

The most consequential legal hearing in AI safety history took place on March 24 before U.S. District Judge Rita Lin in San Francisco. Anthropic sought a preliminary injunction blocking the Trump administration's designation of the company as a national security supply chain risk — a label applied after Anthropic refused to strip safety guardrails from Claude for use in autonomous weapons targeting and mass surveillance programmes. Microsoft and 22 retired military chiefs filed amicus briefs supporting Anthropic's position, arguing that safety-conscious AI suppliers are a strategic asset, not a liability.

The core dispute

Government position: the designation is a contract matter within executive discretion; courts should not second-guess national security procurement decisions
Anthropic's position: the designation is unconstitutional retaliation for exercising First Amendment rights — specifically, publishing an AI safety constitution and refusing to remove it under government pressure
Microsoft's brief: argued that undermining safety-conscious AI companies sets a precedent that deters responsible AI development across the entire US tech sector
Military chiefs' brief: 22 retired generals and admirals argued that AI with intact safety guardrails is more operationally reliable and less exploitable by adversaries than ungoverned AI
Judge Lin reserved judgement; a ruling on the injunction was expected within two weeks

Why this matters beyond Anthropic: The outcome will set a precedent for whether AI companies can maintain published safety commitments under government contract pressure. A ruling against Anthropic could structurally discourage safety investment across the industry.

🧭 iPhone-to-Mac Async Workflow — How Dispatch and Computer Use Fit Together

With the Mac Computer Use research preview now live, coverage on March 24 shifted from announcement to practice — specifically, how to combine Dispatch (released the week prior) with the new computer use capability to build genuine asynchronous iPhone-to-Mac workflows. The pattern: trigger a task from your iPhone via Dispatch, Claude autonomously operates your Mac while you're away, and you return to a completed result. MacRumors and AppleInsider both published detailed walkthroughs of the setup and its current limitations.

The practical workflow

Step 1 — Trigger from iPhone: open Dispatch, describe the task ("refactor the auth module and run tests"), send it to your Claude Code session on your Mac
Step 2 — Claude operates the Mac: Claude uses computer use to open VS Code, edit files, run the terminal, check test output — all without you watching
Step 3 — Get notified: Dispatch sends a completion notification to your iPhone with a summary and any questions Claude needs answered
Step 4 — Review on desktop: sit back down, review the changes, approve or iterate

Best practice for unattended runs: give Claude a clearly scoped task with a defined stopping condition ("stop and notify me if any test fails"). Open-ended tasks risk Claude getting stuck in a loop or making unexpected changes. Use Cowork persistent threads so context carries over if the task spans multiple sessions.

🧭 Anthropic Hits $19B Annualised Revenue — IPO Speculation Intensifies

Research firm Sacra estimated Anthropic reached $19B in annualised revenue by March 2026, up from $14B in February and $9B at the end of 2025 — a trajectory that makes Anthropic one of the fastest-growing software companies ever measured. The figures, extrapolated from API billing and enterprise contract data, fuelled a new wave of IPO speculation following the company's February 2026 funding round at a $380B valuation. Anthropic is now the second-most valuable AI startup globally, behind only OpenAI at approximately $500B.

The numbers in context

$9B ARR — end of 2025
$14B ARR — February 2026 (coinciding with the $30B funding round)
$19B ARR — March 2026 estimate (Sacra)
$380B valuation — implied by February funding round; puts Anthropic alongside SpaceX as an IPO candidate
Growth is driven primarily by API revenue from enterprise customers and the Claude.ai Pro/Max subscription base, not consumer advertising

IPO timeline: Anthropic has not filed publicly, but bankers cited in Fortune expect a 2027 window at the earliest — contingent on the Pentagon legal dispute resolution, continued revenue growth, and broader AI market conditions. A forced IPO timeline due to investor pressure remains a risk given the pace of the funding rounds.