AI company Anthropic plans to grant many more organisations access to its Mythos model. The company announced on Tuesday that approximately 150 organisations across more than fifteen countries will soon use the model.
Around fifty organisations, primarily American companies, initially gained access to Mythos in April. The tool quickly detects security vulnerabilities to help developers improve cyber defences. According to Anthropic, these initial users have already identified at least ten thousand security flaws.
Anthropic, the creator of the Claude chatbot, did not name the specific entities receiving access in this expansion. However, most of the new participants develop software that millions of other organisations, including governments, rely on daily. The company stated that the new cohort includes sectors that lacked representation in the initial phase, such as energy, water, healthcare, communications, and hardware.
Testing and regulatory scrutiny
Major technology corporations including Amazon, Google, Nvidia, Apple, and Microsoft have already used Mythos to test its capabilities. On Monday, a spokesperson for the European Commission confirmed that the European Union Agency for Cybersecurity, ENISA, also received an offer to access the model.
Anthropic intends to roll out Mythos even further in the future. However, the tool is already causing concern among regulators. Watchdogs fear that if Mythos becomes more widely available, criminals could exploit it to launch cyberattacks. Last week, Dutch Central Bank Governor Olaf Sleijpen called AI models like Mythos a gamechanger. He warned that these systems are so fast and smart that they can autonomously execute attacks after finding security vulnerabilities in just a few minutes.
@ anp | NEWS BRAINPORT
