The U.S. government has lifted export controls on Anthropic’s top AI models, Claude Fable5 and Mythos5, raising expectations of renewed competition among companies seeking to deploy AI models in cyber security. OpenAI has also made limited releases of three GPT-5.6 models, including a top model focused on cyber security. Companies are moving to apply agentic AI to security operations centers, while firms and institutions in and outside South Korea outline new security initiatives and strategies.

Akamai Technologies said on Wednesday it is working with Nvidia to introduce a security architecture for AI factories. The companies will integrate Akamai Guardicore Segmentation into Nvidia\'s Vera BlueField-4 STX storage architecture, which runs on the Nvidia DOCA software platform. Akamai said the partnership aims to embed a zero trust architecture as a default layer in AI factories to help enterprises protect data, context memory and autonomous agents.