"""

AI is amplifying existing cybercrime tactics

AI-driven cyber threats are, according to INTERPOL, making established criminal techniques faster and easier to deploy at scale rather than creating a completely new attack model. Speaking at Singapore Tech Week, the agency's global information security chief said fraud and scam operations can now reach far more people at the same time.

INTERPOL Global Information Security Director Bjorn R. Watne said advanced translation tools and the creation of digital identities are making it harder to tell fake interactions from real ones. In his view, the shift is less of a break from the past than an acceleration and expansion of existing tactics.

How should companies set priorities in defense?

Watne stressed that organizations should not build their cybersecurity strategy around trying to respond to every threat at once. Instead, companies should first identify the critical assets essential to keeping operations running, and then determine which actors might try to steal or disrupt them.

Within that framework, threat intelligence is seen as one of the key tools for understanding the tactics and technologies used by attackers. The defensive plan should differ depending on whether it is designed to counter opportunistic criminal groups or more advanced, persistent threat actors.

  • Identify the most important assets needed to keep operations running
  • Determine which threat actors may be interested in those assets
  • Adapt defense layers based on threat intelligence

This approach is also important for directing corporate security budgets and technical resources toward the most sensitive areas.

Warning to boards on new risks from agentic AI

Watne said there is still a gap in many sectors between top management and cybersecurity strategy. He noted that cyber risk is moving higher on corporate risk lists, but in many companies it still does not receive enough attention at board level.

He also warned that extra caution is needed around agentic AI systems, which can act on a user's behalf. Mistakes may go beyond generating incorrect information, he said. As AI is used more often in physical devices such as cars and autonomous vehicles, the risk increases that wrong actions could have real-world consequences.

According to Watne, people tend to be more cautious in financial services, where they are aware of protections such as PIN codes and card-skimming risks, but they place trust in technology products and new apps much more quickly. As AI systems take on more decisions and actions on behalf of users, that level of trust will need to be scrutinized more closely by companies and investors.

"""