Cybercriminals aren’t simply utilizing AI — they’re weaponizing it. Deepfakes, automated phishing, and AI-written malware are rising as a few of the fastest-growing threats on the enterprise radar. In accordance with Foundry’s 2025 Safety Priorities Research, AI-enabled assaults now rank among the many high issues for security consumers, whilst a majority of organizations are investing in or planning to put money into AI-driven defenses. The battle strains are clear: AI versus AI.
Latest CSO reporting paints an unsettling image of what’s already taking place. Autonomous AI brokers are studying to execute full assault chains — from reconnaissance and exploitation to evasion and information theft — with out human path. Researchers have documented AI fashions used to generate extortion emails, launch ransomware, and uncover new vulnerabilities in minutes. As one knowledgeable put it, attackers are “working at pc velocity and scale,” threatening to tilt the steadiness of energy decisively of their favor.
For defenders, the reply isn’t to match automation with automation blindly. Safety leaders interviewed by CSO describe a rising deal with treating AI as a “copilot, not an autopilot.” Nicely-governed AI can speed up detection, triage, and containment, however it nonetheless is dependent upon sturdy human oversight to remain efficient. “The true win isn’t simply velocity,” one CISO advised CSO. “It’s dealing with the routine stuff so analysts can deal with the advanced and strategic issues that machines can’t.”
