DIGITAL LIFE

AI turns phishing into persistent manipulation

Check Point Software Technologies has issued a warning about a new generation of cyberattacks, powered by generative Artificial Intelligence (AI) and autonomous agent systems. These technologies enable the creation of hyper-realistic, multilingual, and adaptive phishing and smishing campaigns, breach traditional defenses, and render human detection virtually irrelevant.

The phenomenon goes beyond ordinary fraud. Cybersecurity experts define this threat as "Advanced Persistent Manipulation" (APM), a direct evolution of APT (Advanced Persistent Threats) that replaces technical infiltration with continuous, automated psychological manipulation.

The era of poorly worded, easily identifiable fraud attempts is over. Generative AI allows attackers to craft flawless messages, based on data collected from social media or data leaks, to create a credible context.

"What were once fraud attempts [...] have evolved into communications virtually indistinguishable from real ones," explains Rui Duro, Country Manager for Check Point Software in Portugal. The company emphasizes that attackers use language models and voice and video cloning tools to "imitate executives and colleagues."

This approach industrializes the exploitation of trust. More worrying is the use of "Agent AI," which coordinates autonomous campaigns. These systems learn from each victim's interaction, autonomously deciding when to switch between email, SMS, or a cloned voice call to maximize effectiveness.

The Threat of Advanced Persistent Manipulation (APM)..."Advanced Persistent Manipulation" is the key concept of this new paradigm. While a traditional APT focuses on technical infiltration to steal data, APM focuses on continuous social engineering.

The goal of AI is not simply to trick the victim into clicking a link. It involves building a relationship of trust over time, manipulating the victim into performing more complex actions, such as approving fraudulent transfers or altering credentials for critical systems.

The scale and zero marginal cost of these attacks place security teams under extreme pressure. Employee training, while necessary, becomes insufficient when the attack is indistinguishable from a legitimate order from a superior.

Reliance on manual processes for analysis and response fails in the face of thousands of simultaneous, personalized attacks. The impact, according to Check Point, transcends cyberspace, compromising the digital trust that underpins operations, finances, and corporate reputation.

The situation is exacerbated by regulatory pressure. Entities such as the SEC and European authorities are demanding greater rigor in incident reporting and risk management. Failure to take measures against AI-based threats can result in sanctions and the loss of insurance coverage.

The answer: fighting AI with AI...Check Point argues that the only viable response is a unified cybersecurity architecture that uses AI itself as a defense mechanism. The company recommends five fundamental pillars, integrated into architectures such as Check Point Infinity:

AI-Powered Defenses: Adopt solutions, such as Check Point Harmony Email & Collaboration, that apply machine learning to detect anomalies and sophisticated phishing attempts that evade traditional filters.

Expanding Zero Trust: Implement strict multi-factor authentication and continuous identity validation across all critical communications, assuming that any request can be compromised.

Multi-Channel Detection (XDR): Use platforms (Infinity XDR) that correlate data from endpoints, networks, cloud, and email to identify the complete attack chain, even if it jumps between different vectors.

Mobile Device Protection: Actively defend against the growing vector of smishing (SMS phishing) on ​​iOS and Android devices, using tools such as Check Point Harmony Mobile.

Automated Response (SOAR): Implement security orchestration to automate incident response, reducing reaction time from hours to seconds. 

Conclusion...The emergence of AI-assisted phishing marks a paradigm shift. Cybercrime no longer relies on individual technical expertise, but rather on access to AI tools that automate mass psychological manipulation. Business resilience requires defenses to evolve from passive detection to active, automated prevention.

mundophone