Security | Threat Detection | Cyberattacks | DevSecOps | Compliance

False positives can disrupt inbound email security as much as missed threats by slowing business workflows and eroding trust in security controls. As phishing attacks become more convincing, many systems respond by tightening filtering thresholds. But without enough context, this can lead to overblocking, where everyday business communication is misclassified as suspicious. Reducing false positives requires more than adjusting filters.

With over 10 years of experience in implementing AI, KnowBe4 has a ton of agents on its platform which customers can use to significantly lower risk. They help to secure the digital workforce (humans + AI agents). But five of them, all based and driven by risk scoring metrics, have become my favorites.

Cyber insurance claims surged by 40% over the past eighteen months, while ransomware payments have dropped by 44%, according to a new report from Cowbell Cyber. The three most common incident types were data breaches, cybercrime (including phishing and business email compromise), and extortion attacks (including ransomware).

Researchers at Push Security have analyzed a phishing platform used by organized criminal threat actors like ShinyHunters and BlackFile, finding more than 400 domains linked to attacks launched by the phishing kit.

The classic 'survey reward' scam is back and hitting harder than ever. KnowBe4 Threat Labs is tracking a massive, high-volume campaign that is not only impersonating a wide array of trusted global brands across retail, logistics, and healthcare, but is using hundreds of newly registered domains (NRDs) and sophisticated psychological priming to fly past traditional security defenses.

In 2024, we talked to AI. In 2026, AI is talking to our systems, our customers, and increasingly, acting on our behalf. With AI agents, we are moving AI from a tool to an actor, from assistance to agency and from outputs to actions. And that changes the nature of risk. AI agents plan, execute, and interact with the world on our behalf. They send emails, move data, trigger workflows, and increasingly operate across systems without human intervention.

Threat actors are increasingly augmenting their attacks with AI tools, according to researchers at Google’s Threat Intelligence Group (GTIG). For the first time, GTIG observed a threat actor using a zero-day exploit developed by AI, although Google blocked the attack before it succeeded. Threat actors also continue to use Large Language Models (LLMs) for research, reconnaissance, and malware development.

The era of "typing into a box" is over. For years, we viewed artificial intelligence as a digital assistant—a sophisticated autocomplete tool that waited for human input. But according to Martin Kraemer, KnowBe4’s CISO Advisor for Europe and the Middle East, that dynamic has shifted. We have moved from asking AI questions to giving AI jobs. In a recent deep-dive webinar, Martin explored the transition from AI tools to AI agents.

2026 has officially become the year of speed, scale and support. The delta between a phishing email landing and a full organizational compromise has shrunk to mere seconds. The reality by the numbers: To close this window, your defense strategy must evolve into a two-step strategy of accuracy and automation.

There is no question that AI is changing cybersecurity in a massive way. In many respects, its impact is comparable to the rise of the internet. AI tools are helping organizations improve efficiency, automate repetitive tasks, and process data at a speed humans simply cannot match. Unfortunately, the same technology helping defenders is also being adopted by cybercriminals just as quickly. For cybersecurity professionals, keeping up with AI and agentic developments is no longer optional.