The security solutions provider company Hornetsecurity has identified the main cybersecurity trends that will mark the daily lives of companies in 2026, with special focus on cyberweapons based on agentic AI.

The Security Lab team has been able to review and analyze more than 72 billion emails and all this information has helped the company better understand the situation in the sector and study attack trends that can help businesses anticipate threats, highlighted Daniel Hofmann, CEO of Hornetsecurity.

Top security threats in 2026

Among the main trends in cybersecurity for 2026 highlighted by Hornetsecurity’s “Cibersecurity Report”, the following stand out:

Uncontrolled adoption of AI tools

The adoption of AI in companies is accelerating, both through management initiatives and informal experiments by employees themselves. This pace of innovation will outpace the ability of legal, IT, and security teams to evaluate each use case, expanding the attack surface across enterprises.

Cyber ​​weapons based on agentic AI

Agentic AI systems are already being used as cyber weapons. Criminals will be able to write scripts, adapt them in real time and launch complex campaigns with very little technical knowledge. These models will allow you to automatically generate phishing baits, overcome CAPTCHA systems and impersonate identities with voice and video deepfakes.

Ransomware 3.0 irruption

The use of ransomware is entering a new phase, marked by data manipulation. Historically, Ransomware has evolved in response to defensive improvements: Ransomware 1.0 (attacks focused only on encryption) and Ransomware 2.0 (double extortion, with encryption and data exfiltration). In 2026, Ransomware 3.0 will no longer only allow data to be encrypted or stolen, but will also alter its integrity, modifying or corrupting records in a subtle way, with the aim of raising doubts about the reliability of the information. Manipulating data in financial systems, medical records or industrial controls could cause significant damage.

Quantum computing

Cryptographically relevant quantum computers (CRQC) could still take 5-15 years to become available. These computers will not replace current ones, but will be used to make very specific calculations, such as breaking current encryption algorithms. The problem is that, even if they haven’t arrived yet, companies must start preparing, especially if they manage personally identifiable data (PII) and health information (PHI) and plan to retain it (or are required by law) for more than five years.

In those cases, quantum computing-resistant algorithms are needed to encrypt that data as soon as possible. The reason: that some criminals already apply the “Harvest Now, Decrypt Later (HDNL)” strategy, which consists of collecting data, even if they cannot yet decrypt it, with the intention of doing so in the future, when they have quantum computers capable of breaking the current encryption.

The new Proofpoint after the acquisition

Just a month ago, Hornetsecurity Group was fully integrated into Proofpoint, once the acquisition process of the company specialized in cybersecurity was concluded. The transaction represents an important milestone in Proofpoint’s strategy to extend its people-centric security platform to organizations of all sizes through a specific approach to the MSP channel.

Thanks to Hornetsecurity’s consolidated presence in the European market and its partner network that serves more than 125,000 customers through more than 12,000 MSPs and channel partners, Proofpoint considerably increases its capillarity and capabilities in the SME segment to combat cyberweapons based on agentic AI.

Hornetsecurity’s flagship solution, 365 Total Protection, is a cloud security platform for M365 environments. With services spanning email security, backup, compliance, awareness training and access control, the platform offers MSPs a multi-tenant control panel to manage and scale their clients’ protection, without the complexity of integrating multiple solutions.