Channel Insider content and product recommendations are
editorially independent. We may make money when you click on links
to our partners.
Learn More
At Proofpoint Protect 2025 in Nashville, Tennessee, AI in human-centric security was a major point of emphasis for the enterprise cybersecurity solutions provider.
How Proofpoint leverages AI to drive security outcomes forward
The company leverages AI substantially for threat detection, while also utilizing Proofpoint Satori Agents to support SOC teams as a force multiplier in managing resilience.
Proofpoint’s detection approach focuses on eliminating known threats and progressively analyzing emails using AI models.
One of the consistent messages during the event was that “every email is now a prompt.” Meaning that every email received by an organization can feature prompt injection for use in phishing scams.
Satori Agents, introduced by Proofpoint, handle data loss prevention alerts, automate phishing simulations, and resolve user-reported email threats for enterprises. Proofpoint’s Satori DLP Triage Agent, Satori Phishing Simulation Agent, and Satori Abuse Mailbox Agent help to secure the agentic workspace in the age of AI.
“As we progress, as we gain more trust and are able to build more guardrails and capabilities into the system that will graduate from a supervised agent to something that we characterize as having bounded autonomy, and then finally to fully autonomous agents,” said Chief AI & Data Officer at Proofpoint, Daniel Rapp. “But we’re taking a measured approach to this initially. You can have a prompt in the text of your email. In the world of AI, any sort of media that we act with as a human being can be a payload.”
The threat actors are using AI, too
What’s more is that threat actors are becoming more patient, and AI is enabling them to have the time to get more creative.
Michael Frendo, CTO of Engineering at Proofpoint, highlighted that threat actors are targeting various communication vectors beyond email, including WhatsApp, SMS, and Slack.
“It’s all about different kinds of interactions, and protecting all of those interactions because they’re all vulnerable in the sense that they create an attack surface,” said Frendo. “You can deliver a URL, for example, through text message or through WhatsApp that we need to protect in that direction as well. We have a database of over two trillion URLs that we can mark as either safe or not safe, and we use that information with our collaboration protection.”
Partnerships and integration with others also central to strategy
Rapp mentioned that partnerships with Microsoft and CrowdStrike to integrate agentic AI capabilities into their respective products has been key to enhancing Proofpoint’s offerings.
The collaborations aim to provide a stronger integration between Proofpoint’s agentic AI and other vendors’ security tools. A team approach is essential for cybersecurity, the Proofpoint execs said, and that no single vendor can address all aspects of security.
“We were at CrowdStrike Fal.Con last week, where we announced the partnership to provide a stronger integration between Charlotte AI and the Fusion SOAR that CrowdStrike has with using our agentic AI capabilities, we’re really excited about this,” said Rapp. “We think it’s going to open up some incredible opportunities because cybersecurity is not going to be won by any one vendor. It really is a team sport.”
Frendo adds that over the next 18 months, they expect there to be more interactions amongst agentic processes than there are amongst people today.
He relates it to when a camera was put on a phone, and the number of pictures taken in a year exploded beyond the number of images taken in the history of photography.
“This is going to be a similar dynamic, right? There’s going to be billions of agentic systems out there that are interacting with each other, and they interact much faster than people do,” Frendo explained. “The projection is that the number of interactions between agentic processes, between agentic processes in people, and agentic processes in each other will actually supersede the human interactions in the next 18 months, and we’re really just getting started from a real perspective.”
Frendo adds that while the exponential growth of interactions will be a real challenge, it will also present a significant opportunity for organizations.
Proofpoint’s four innovations announced during Proofpoint Protect 2025 are meant to secure AI workspaces, protect assistants, govern agents, and safeguard data. Read more about those announcements and when these agents will be available.
