Palo Alto Networks tested Anthropic’s Claude Mythos 5 model on its internal infrastructure in April. In roughly one month, the AI discovered 75 vulnerabilities across more than 130 products, while the company normally finds fewer than five per month. The model successfully generated working exploits more than 70 percent of the time. The US government imposed export controls on the model around June 12, partially lifting them on June 26. Palo Alto Networks launched its Unit 42 Continuous Frontier AI Defense service on September 22.
Source: Read the original article
Disclaimer: this content is for information purposes only and is not financial advice. Cryptocurrencies are highly volatile: you may lose all of your capital. Always do your own research. Legal notice
