Analyst firm Gartner has identified autonomous artificial-intelligence agents for security operations centers as the “next big innovation” in enterprise defense, saying the software can investigate incidents and weed out false positives in minutes, according to a new briefing highlighted on 13 August. The assessment coincides with discussions at the BlackHat 2025 conference in Las Vegas, where industry executives warned that generative AI is enabling faster, more convincing phishing campaigns and compressing the time between vulnerability discovery and exploitation. Speakers said security teams will have to adopt real-time, machine-driven countermeasures to keep pace, while retaining human analysts for the most complex decisions. Researchers also stressed the need for systematic “red-teaming” to expose weaknesses in defensive AI models before criminal hackers do. A series of online workshops curated by cryptographer Bruce Schneier, branded “Prompt||GTFO,” is sharing early field results and practical techniques, with the fourth session scheduled later this year.