Anthropic, the company behind the Claude AI assistant, has broadened the availability of its advanced cybersecurity-focused model to organizations in over 15 more countries. This expansion is part of Project Glasswing, the company’s initiative providing selected entities access to Claude Mythos Preview.
The newly included nations feature Australia, Canada, France, Germany, Italy, Switzerland, the Netherlands, Spain, Belgium, Sweden, India, Japan, New Zealand, and South Korea. This geographic expansion adds approximately 150 new organizations to the program, provided they meet Anthropic’s stringent security requirements.
Project Glasswing initially launched in April with a limited cohort of around 50 companies. Since then, these partners have used the Mythos model to scan their software, collectively discovering more than 10,000 high- and critical-severity security flaws.
The expanded group now includes industries previously underrepresented, such as power, water, healthcare, communications, and hardware. Many of the new participants are vendors or nonprofits that maintain codebases relied upon by numerous other organizations worldwide, including governments.
“What each partner has in common is that a successful attack on their codebase could be catastrophic,” Anthropic stated. The company added that a major cyberattack on most participants could potentially impact more than 100 million people.
Despite the expansion, Anthropic reaffirmed it has no plans to release Mythos-level capabilities for general public use. The company cited the absence of strong safeguards against misuse as the primary reason for this restriction.
Anthropic warned that similar advanced cyber-capable AI models could become widely available from other companies within six to 12 months, potentially without adequate safety measures. The startup believes such models, if used maliciously, could make cyberattacks more frequent and unpredictable.
Consequently, Anthropic is maintaining controlled access for trusted organizations while it continues to develop stronger safety protections for this powerful technology.
This news article is created with AI enabled services and auto published, therefore it might contain errors, reader discrimination advised.
This article and views expressed therein, are solely that of the original news source also displayed for the interested reader, therefore it is not an act of copyright infringement by this platform, also this platform only aggregates the content and has no bearings and or shares any opinions or promotes views expressed by the original author, all intellectual property rights is totally credited to the original source.
