OpenAI and Anthropic Cyber War

The release of GPT-5.4-Cyber marks a significant escalation in the competition between the world's leading AI labs.

OpenAI claims its new model is designed for defensive tasks, though experts warn that such capabilities could be dual-use for offensive hacking.

Anthropic's Mythos has already drawn concern from the U.S. Treasury, which is seeking access to the tool to hunt for flaws in financial systems.

This 'cyber cold war' highlights the growing tension between rapid AI innovation and national security requirements.