Security | Threat Detection | Cyberattacks | DevSecOps | Compliance

2025 was a turning point for identity security. Many professionals realized that traditional human and machine-focused identity solutions just don’t work for AI. AI is non-deterministic like a human, yet it’s still software. This creates an entirely new identity category. Traditional IAM tools would treat AI identities as yet another separate type, creating new silos.

APIs now sit at the center of almost every digital product, from mobile apps and SaaS platforms to embedded services. As organizations scale, the number of endpoints grows quickly, as does the attack surface. Unmonitored or misconfigured APIs have already led to major incidents across industries, including data exposure, broken authentication, and large-scale account takeover.

I built a demo showing how to wire up multiple AI agents using Google's Agent Development Kit (ADK) and the A2A protocol, with GitGuardian scanning content for secrets.

News AI Usage AI Risk Ria Manzanero Head of Marketing 18 December 2025 3 min read Share.

Looking back at 2025, two mega-trends from the past have continued: First, data breaches remained a constant and continued to trend upward; and second, there was once again a headline disaster no one anticipated. The first point needs no elaboration; data breaches are like air pollution—an accepted nuisance that only occasionally becomes so severe that we wonder why we live like this. For the second point, I gesture toward the major incidents of recent years. MoveIt. Crowdstrike. Snowflake.