Cardano founder Charles Hoskinson criticised Ethereum co-founder Vitalik Buterin’s warning that advances in AI-driven mathematics could weaken lattice cryptography within two years. Hoskinson called the claim unfounded fear and said it relies too heavily on speculation without mathematical proof. He said lattice cryptography has been studied and rigorously tested for decades, with security assumptions refined over time. He also warned the view could harm cybersecurity efforts to protect internet communications from quantum computers.

Security companies are stepping up AI security while expanding beyond traditional security into AI governance, risk control and cost management. PRIBIT AI is shifting its identity from a security company to an AI governance and risk-control firm and has announced an execution strategy. F5 has introduced an AI gateway that applies consistent policies to AI requests and manages access and usage for models, agents and tools. Korean companies are also joining OpenAI-led cybersecurity cooperation initiatives.