Growing AI adoption may expose PHL companies to more cyberattacks

PHILIPPINE BUSINESSES may face more cyberattacks as they continue to adopt artificial intelligence (AI)-powered solutions, with threat actors also using more advanced tools to exploit these technologies, cybersecurity company Trend Micro said.

Trend Micro Philippines Country Manager Ian Felipe said at a briefing on Monday that Philippine enterprises have improved their overall cyber hygiene, boosting investments to address cybersecurity risks and improving their teams’ skill sets.

However, firms must be aware of the risks that come with using AI-driven technologies, he said.

“It’s very important to have visibility in terms of what might and what can happen in the future when you adopt AI,” Mr. Felipe said.

Trend Micro said cyberattacks are now targeting AI agents, with criminals hijacking or misguiding these agents to exploit firms’ systems.

“When you use agentive AI, it’s basically an autonomous AI that can decide on its own, so having this kind of setup provides less visibility in terms of the actions or interactions of this AI to our systems,” Raymond Almanon, senior threat researcher at Trend Micro, said at the same briefing.

Agentive AI deploys a self-governing “agent” that can make decisions without minimal human supervision. It is expected to help automate tasks, reduce costs, and improve productivity.

According to Trend Micro, attackers are also using AI to launch faster and undetectable attacks with less tools.

“Adding to this is the unauthorized or malicious activities carried out by misguided autonomous agents, so we really need to make sure that every time we deploy an agent, there is a way for us to monitor it,” Mr. Almanon said.

He said businesses should have measures to monitor their AI agent operators to spot potential vulnerabilities.

“We have to implement robust security measures that give end-to-end visibility on agent operations and harness AI to ensure protection from vulnerabilities it creates for itself.”

With more companies using AI technologies, hackers are also likely to target AI-driven training environments, access infrastructure, and service subscriptions to steal information, Mr. Almanon added.

Many cyberattacks focus on obtaining companies’ login credentials to help them orchestrate large-scale exploits, he said. “Once info stealers get access to one of your employees’ credentials, that can be the start of a bigger attack on your network.”

AI can also be exploited to craft phishing e-mails, conduct disinformation campaigns and scams like pig butchering or investment fraud, expand the scale of cyberattacks, and target device drivers, he said.

“If you have a vulnerable driver that is installed — meaning these are pre-installed on your machine or came from another software — they can be targeted and used to execute malicious codes without you knowing,” Mr. Almanon said. — Beatriz Marie D. Cruz