Skip to content
Theme
OnDevice Tools

GadgetsLaptops & PCs

Agentic AI cyber threats force new security urgency

Proofpoint says AI-driven attacks are speeding up and forcing companies to rethink access, data protection and response.

Illustration of a laptop showing AI security warnings and a protected company network

Key takeaways

  • Proofpoint chief executive Sumit Dhawan warned that AI agents can act like insiders inside company systems.
  • Proofpoint says automated AI tools can find flaws, launch attacks and create deceptive campaigns at speed.
  • The company is promoting Defend with Intent and Access with Intent to cover people, data and AI usage.
  • Smaller businesses are urged to go beyond antivirus, firewalls and spam filtering as AI-enabled threats grow.
In this article
  1. What Proofpoint warned
  2. Why attacks are moving faster
  3. Defend and access with intent
  4. Smaller businesses under pressure
  5. What defenders should do next

What Proofpoint warned

According to TechRadar, Proofpoint chief executive Sumit Dhawan has warned that agentic AI cyber threats are creating a new wave of attacks. He spoke at the Proofpoint Protect Tour London 2026 event. His message was that companies of all sizes need to treat AI agents as insiders, because these automated tools can act inside business environments.

Proofpoint says its research shows criminals can use AI to operate at a larger scale than before. The company says attacks are not limited to network intrusions. They can include automated probing, rapid social manipulation and actions across many targets.

The warning places AI at the centre of modern security planning. Instead of seeing AI only as a defence tool, Proofpoint says businesses must also manage the risks created by AI use. That includes watching how people and automated agents interact with company data.

Why attacks are moving faster

Illustration: warning, security, AI, building and the labels Agentic AI, Hornetsecurity, Proofpoint

Dhawan said the time defenders have to react is getting shorter. He pointed to a cyber theft that took only 27 seconds. He used the example to show how automation can turn a small opening into a fast incident.

He said agentic toolkits can find weaknesses, launch attacks and produce deceptive campaigns automatically. These campaigns can target employees and customers through messages designed to trick them. Because attackers can reach people through many channels, Proofpoint says reachability has become a major security concern.

The company argues that defence must be based on intent. That means security teams need to understand not just what an action looks like, but why it is happening and what it is trying to reach. This is especially important when automated systems can move faster than human review.

Defend and access with intent

Proofpoint is promoting two approaches: Defend with Intent and Access with Intent. The first focuses on learning where attacks begin and improving resilience. The second focuses on the way AI agents use data and applications, which can differ from normal human behaviour.

Dhawan said security has little value if it does not cover the whole environment. He said a typical company could see vast numbers of AI prompts and actions each year. Reviewing all of that activity manually would be impractical, so tools need to watch behaviour and context.

Proofpoint says its strategy is to protect behaviour and keep data connected when people and AI work together. The company also says integration matters, because customers need protection that fits across their existing systems. It points to acquisitions such as Hornetsecurity and Varonis as part of this effort.

Smaller businesses under pressure

The warning may be especially important for small and medium businesses. Dhawan said these organisations once relied on basic tools such as antivirus, firewalls and spam filtering. He said that approach is no longer enough in an environment where AI-enabled attacks can be fast and convincing.

Daniel Blackford, who leads threat research at Proofpoint, said smaller firms should treat security as a higher priority. He warned that an attack could damage reputation. For smaller firms, damage to reputation can be especially serious.

Proofpoint says its acquisition of Hornetsecurity is aimed at giving smaller businesses stronger protection. Its acquisition of Varonis is presented as part of a broader push into AI-related security. Together, the moves show how the company is trying to cover both human-focused threats and data-focused risks.

What defenders should do next

Blackford said attackers have adopted AI in much the same way as defenders. He noted that both sides are still learning how to use the technology effectively. However, he said the period in which defenders held an advantage is narrowing.

Dhawan said the key issue is not whether AI security is overhyped. The key issue is what organisations do in response. He said security teams face a difficult task because they are always chasing major threats, and constant rapid reaction can become exhausting.

His advice was for every organisation to invest in the right setup for access and data. That means making sure access rules and data controls are designed for an environment where humans and AI agents both take actions. Proofpoint says this is necessary to reduce the risk of a new wave of cyber threats.

Sources

  1. 'Every organization has to invest in getting this right': Proofpoint CEO warns the agentic AI age will mean a whole new generation of cyber threatsTechRadar

Tags

Spotted a mistake? See our corrections policy. Read how News by OnDevice works in our editorial policy.