Anthropic

$16.3B in H1 2026 revenue and an $80.7B net loss put Anthropic at the center of an AI race increasingly defined by safety, government access and compute intensity.

Who they are

Anthropic is an AI company whose coverage centers on its Claude models and applications, its safety-led positioning under CEO Dario Amodei, and its role as a frontier-model rival to OpenAI. It appears both as a commercial model provider and as a consequential counterparty to US government and national-security debates over how advanced AI may be deployed.

The recent arc

Coverage accelerated sharply through 2026, peaking in Q3, after shifting from product competition into an extended dispute with the Pentagon and Trump administration. In late February and March, Amodei said Anthropic could not remove safeguards at the DOD’s request and would challenge its risk designation; reporting on the failed talks said the Pentagon sought AI analysis of bulk data collected about Americans. OpenAI then explicitly contrasted its own DOD agreement with Anthropic’s in a March 1 statement.

The recent arc

June made the confrontation operational: Anthropic introduced Claude Fable 5 and restricted-access Mythos 5, then said it was disabling both after an export-control order, while Axios and Politico reported attempts to resolve the White House dispute. The latest coverage broadens the frame again, pairing the Claude Duet proactive-agent preview and a protein-minibinder project with Citrini Research with IPO-prospectus disclosures of rapid revenue growth, high infrastructure spending and customer concentration.

The tension

The central tension is whether Anthropic’s safety constraints are a competitive advantage or a constraint in the race for government adoption and frontier-model scale. Amodei frames a costly “race to the top” on safety, while Trump, David Sacks and parts of the administration challenged the company’s approach during the Mythos 5 conflict; OpenAI is the clearest rival beneficiary and comparator, particularly on classified and DOD deployments.

Why it matters

If this trajectory holds, Anthropic’s significance will rest on whether it can turn Claude’s safety distinction and growing product reach into durable enterprise and government-scale demand while funding the compute required to compete. The draft S-1 figures make that balance unusually visible: growth is substantial, but so are infrastructure demands, accounting-driven losses and reliance on two customers, leaving safety policy, public-market scrutiny and supply of computing power tightly linked.

Related: Claude · OpenAI · Dario Amodei · Trump · Dario Amodei says Anthropic cannot “in good conscience” accede to DOD' · Sources: senior Anthropic technical staff are in DC to meet WH officia