AI-Driven Cybercrime Consolidation

He also explained the staggering projected economic impact of this criminal industry, which is estimated to exceed one trillion US dollars by 2025, a figure that dwarfs the legal cybersecurity insurance market, which is valued at less than USD 15 billion.

According to the LACNIC expert, AI has become an exponential driver in the escalation of cyber threats. Old tactics such as social engineering, phishing, and credential theft are being enhanced with AI to create attacks that are more convincing and harder to detect.

Adding to this are video and audio deepfakes, which represent a new level of extortion and fraud. The rapid adoption of this technology as a weapon is reflected in the 89% increase in attacks involving the use of AI across the region in just one year.

Carlos Martínez observed that emerging trends are creating an increasingly broad attack surface. The Ransomware-as-a-Service model has become commoditized, while exploits and attacks are increasingly focused on cloud, SaaS, and API environments, which are often among the least protected.

He added that attacker speed is critical, as data from Crowdstrike shows that by 2025, the average time for an attacker to make the first move on a minor target was only 29 minutes, and 32% of zero-day vulnerabilities were exploited before or on the same day they were disclosed.

Given this scenario, the concept of a “security perimeter” is considered outdated. Remote work, VPNs, and the extensive use of the cloud have expanded and fragmented the attack surface.

Additionally, effective protection can no longer rely exclusively on perimeter firewalls, as the new paradigm demands that organizations build defenses at each point of interaction—in identity verification, in every protected service, and in every authorized access—completely redefining their digital defense strategy.