Artificial intelligence is helping criminals execute familiar scams faster and target more victims, INTERPOL’s global chief information security officer Bjorn R. Watne said at Tech Week Singapore on Thursday. He urged companies to identify their most critical assets, determine which adversaries might target them, and tailor defenses using threat intelligence rather than attempting to protect against every possible threat. Improved translation and digital identities can make fraudulent communications harder to distinguish from legitimate contacts. Watne said opportunistic criminals and advanced persistent threat (APT) actors require different defenses, while cybersecurity remains insufficiently represented in many boardrooms despite rising on corporate risk registers. TeamT5 found Chinese state-affiliated groups more than doubled attack volume after adopting open-source AI models. Watne also warned about agentic AI, or systems that act for users, particularly in cars and self-driving vehicles, where incorrect actions could cause bodily harm. CISA issued joint guidance in May urging organizations to assume agentic AI may behave unexpectedly until security standards mature. OpenAI warned in August that AI-enabled attacks would become far more widespread within months; more than 100 organizations, including Anthropic, Google, Microsoft, Coinbase, and Circle, had signed its open letter at launch.