Brief IA
›

Proofpoint Calls for a Review of AI Agent Access

Business & Startups·Tom Levy·

Proofpoint Calls for a Review of AI Agent Access

Proofpoint Calls for a Review of AI Agent Access
Key Takeaways
1According to Sumit Dhawan, an average company can generate millions of AI prompts and actions per year.
2Daniel Blackford believes that the lead of advocates in the use of AI is narrowing.
3Proofpoint recently acquired Hornetsecurity to enhance protection for SMEs.
Why it matters : Proofpoint believes that securing access for AI agents requires a different approach than that applied to humans.

For Sumit Dhawan, CEO of Proofpoint, companies must frame their AI agents as internal stakeholders. At the Proofpoint Protect Tour London 2026, the leader called for investment in data and access management in the face of increasing AI-driven attacks.

SMEs Must Go Beyond Antivirus, Says Dhawan

According to Sumit Dhawan, an average company can generate millions of AI prompts and operations over the course of a year, a volume that cannot be fully controlled manually. He also anticipates that the need to constantly and quickly respond to threats could exhaust employees.

For SMEs, Dhawan and Daniel Blackford recommend strengthening their security practices. Dhawan believes that antivirus, firewalls, and spam filtering are no longer sufficient. Blackford emphasizes the growing importance of cybersecurity for these businesses, highlighting that an attack could severely damage their reputation. Proofpoint recently acquired Hornetsecurity to provide comprehensive protection for SMEs.

Blackford Sees Defenders' Lead Diminishing

Daniel Blackford, head of threat research at Proofpoint, observes that the advantage defenders had in using AI is diminishing. In a media exchange alongside Dhawan, he explained that attackers are following a similar trajectory of AI adoption as other users, and both sides are still figuring out how to best leverage these tools.

Proofpoint indicates that its recent research shows that AI enables cybercriminals to operate on an unprecedented scale. Dhawan cites agentic tools capable of identifying vulnerabilities, launching attacks, and automatically generating multiple social engineering campaigns. He also mentioned a cyber theft carried out in 27 seconds to illustrate the increasing speed of attacks.

Le brief IA que lisent les pros

Le brief IA que les pros lisent chaque soir

Les 7 actus IA du jour, décryptées en 5 min. Gratuit.

Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.

Choisissez votre rythme

Gratuit · Pas de spam · Désabonnement en 1 clic

Dhawan prioritizes the question of AI usage in business over concerns about its potential transience. He also advocates for the combination of AI and community efforts to counter attackers.

Proofpoint Aims to Adapt Protection to Agent Access

The warning issued in London by Sumit Dhawan concerns companies of all sizes. He recommends considering AI agents as internal actors and calls for investment in configuring access and data. According to him, agents do not access applications and data like humans do, which necessitates a different security approach.

Proofpoint structures its strategy around "Defend with Intent" and "Access with Intent." The first aspect aims to understand the origins of attacks and organize defense and resilience, particularly against tools provided by AI. The second seeks to ensure effective collaboration between humans and AI agents. Dhawan emphasizes the protection of individuals, access management, and understanding intentions.

The company, which aims to protect its clients against as many threats as possible, recently acquired Varonis as part of its AI-related protection ambitions. For Dhawan, protection must be compatible with all elements of the client's IT environment. He presents integration as a continuous goal and states that Proofpoint is working to secure behaviors and maintain data connectivity in environments where humans and AI collaborate.

Brief IA — L'actualité IA en français

L'essentiel de l'actualité de l'intelligence artificielle, décrypté et expliqué chaque jour.