AI & Enterprise
OpenAI says enterprise security must shift to agent-centric model in AI era
OpenAI warned that open-weight AI models have relatively low guardrails and, after fine-tuning and additional computation, can gain cyberattack capabilities approaching frontier models, requiring safeguards against misuse. Wana Tun, an applied AI security specialist at OpenAI, said malicious actors could fine-tune such models to cause significant harm. She said AI-driven cyberattacks are becoming more autonomous and called for enterprises to shift defenses toward AI agents to speed detection, verification and patch creation.
