Spain Gets Early Access to Anthropic’s Mythos AI
What is Mythos?
Mythos is a new artificial intelligence model created by the startup Anthropic. It is designed to scan computer systems and find weak spots very quickly. The model can also simulate sophisticated cyber‑attacks, which makes it a powerful tool for both defense and offense.
Why the Model Raised Alerts
When Anthropic first announced Mythos in April, experts warned that its ability to uncover vulnerabilities could be misused. Because of these concerns, the company decided not to release the model publicly. Instead, it offered access only to a limited group of trusted partners.
Opening the Door to Europe
Interest in Mythos spread quickly, and many organizations asked for a chance to test it. The European Union, through its cybersecurity agency ENISA, pressed for early access. Anthropic responded by inviting ENISA and a select set of 150 institutions from over 15 countries to join a program called “Project Glasswing.” Participants receive a preview version of Mythos to study its risks and benefits.
Spain’s Role in the Program
Following the EU’s lead, the Spanish government requested early access to Mythos so the country would not fall behind in AI developments. In early May, Spain’s Vice President and Minister of Economy, Carlos Body, asked for the opportunity to test the model. The request was granted, and Spain is now part of the testing group.
How Spain Plans to Use Mythos
Spanish officials say they will use the preview version for defensive purposes only. By running Mythos against their own networks, they hope to discover weaknesses before hackers can exploit them. Government sources describe the move as a sign of Spain’s growing influence in global AI governance.
Anthropic vs. OpenAI
While Mythos makes headlines, Anthropic’s other AI product, Claude, is also gaining traction. Claude is a generative AI model that competes directly with OpenAI’s ChatGPT. Recent reports show Anthropic attracting more business customers than OpenAI and projecting a profit of about $559 million for the next quarter. In contrast, OpenAI expects to lose billions each year.
What Experts Think
Some security specialists remain skeptical about Anthropic’s claims, saying the company’s statements need more verification. Independent evaluations, however, suggest that Mythos is indeed a potent offensive tool, which is why access is tightly controlled.
Conclusion
Spain’s early access to Mythos highlights how countries are racing to understand and control powerful AI technologies. By testing the model in a secure environment, Spain aims to strengthen its cyber defenses while contributing to the broader conversation about responsible AI use. The situation also shows the growing competition between AI firms like Anthropic and OpenAI, shaping the future of technology worldwide.
Images Credit: www.diariodemallorca.es