Anthropic Expands Mythos AI Model to 150 New Partners Across 15 Countries

CNBC reported Tuesday that Anthropic is broadening Project Glasswing to encompass 150 additional partner organizations spread across more than 15 countries. The move scales up the company’s controlled deployment of its Mythos AI model, which is specifically designed to detect software vulnerabilities before bad actors can exploit them.

New Industries Enter the Glasswing Program

The latest wave of partners extends into sectors that were absent from the program’s initial launch. Power utilities, water infrastructure operators, healthcare providers, communications firms, and hardware manufacturers are among those gaining access. Each incoming partner must satisfy a set of security requirements before Anthropic grants them access to Mythos.

Anthropic described the expansion as a step toward a broader ambition. The company said in a blog post that its long-term goal is for AI to make all software more secure and to help the industry adapt to the ways AI could reshape core cybersecurity assumptions.

Also Read: OpenAI Launches GPT-5 With Advanced Reasoning Capabilities

Background: A Cautious Launch Built Around Risk

Project Glasswing began in April with a deliberately limited cohort of roughly 50 testing partners. The restrained rollout reflected genuine concern inside Anthropic about Mythos’s raw capability. The model’s ability to surface software weaknesses raised alarms that it could equally be weaponized by malicious actors to probe systems for entry points.

Those concerns prompted discussions involving the White House, major technology companies, and financial institutions. Since that initial April launch, partners have collectively surfaced more than 10,000 security flaws rated as high or critical severity, suggesting the controlled program has already delivered measurable defensive value.

Also Read: Anthropic Raises $2.5 Billion in Latest Funding Round

Glasswing Expansion Follows a Busy Week for Anthropic

The Mythos announcement lands amid a flurry of corporate activity at the San Francisco startup. One day before Tuesday’s expansion news, Anthropic confirmed it had confidentially submitted its IPO prospectus to the SEC, moving ahead of rival OpenAI in that particular race. The company also began offering Mythos access to partners within the European Union on Monday, signaling a parallel push into regulated international markets.

Together, the moves paint a picture of a company accelerating on multiple fronts simultaneously, from commercializing its most sensitive AI tools to preparing for life as a public company.

Read Next: Anthropic Files Confidential IPO Prospectus With the SEC