NewsNTech
The cost of running AI at security-operations scale is the constraint that shapes enterprise buying decisions.
Microsoft is now claiming its new cybersecurity model addresses that pressure directly, and says the system, when integrated with OpenAI's GPT-5.4, can outperform Anthropic's new Mythos 5.
Microsoft frames the model as cost-saving, a description aimed at security teams weighing the per-query economics of large-model deployments.
The competitive claim zeroes in on Mythos 5, Anthropic's newest entry in the space. Microsoft says the combination with GPT-5.4 clears the performance bar against that target.
Keep reading