Proofpoint has unveiled four new technologies to protect what it calls the “agentic workspace”, where humans and AI agents work side by side. The cybersecurity and compliance company says the solutions help organisations embrace artificial intelligence safely by defending against new threats, preventing data loss and improving oversight of AI assistants.
As AI agents become more common in workplaces, they are used to automate tasks, analyse data and collaborate with people and other AI systems. But these agents behave like humans — clicking links, sharing files and taking action — making them vulnerable to manipulation, data theft and social engineering attacks.
“The agentic workspace is here and one of the most profound shifts in terms of how work gets done,” said Sumit Dhawan, chief executive of Proofpoint. “Protecting the agentic workspace is the next evolution of human-centric security, extending beyond people to safeguard AI agents and the points where they collaborate and share data.”
Four innovations to secure collaboration and data
The first solution, delivered through Proofpoint Prime Threat Protection, blocks malicious prompts hidden in emails. These attacks target AI assistants like Microsoft Copilot and Google Gemini, tricking them into leaking sensitive data or bypassing defences.
Proofpoint Data Security Complete helps organisations find and classify sensitive data, control access and monitor how it is used across endpoints, email, the web and cloud applications. It integrates several tools — including data loss prevention, insider threat management and data lineage tracking — into one platform. Proofpoint AI Data Governance adds oversight of AI usage, helping teams detect unauthorised tools, enforce security rules and automate access controls.
Proofpoint Secure Agent Gateway focuses on protecting AI agents deployed by companies. Using the Model Context Protocol, it controls how these agents handle data, monitors their activity and blocks or redacts sensitive details before information is shared further.
Finally, Proofpoint Satori Agents work inside the company’s platform to automate routine security tasks such as responding to phishing reports or managing data alerts. Satori MCP Access lets external AI tools like Microsoft Copilot or CrowdStrike Charlotte interact with Satori Agents, speeding up threat detection and response.
Driving safer AI adoption across industries
Ryan Kalember, Proofpoint’s chief strategy officer, said the rise of AI is forcing organisations to rethink cybersecurity. “By uniting collaboration and data security, we are redefining how organisations secure work in the age of AI. Proofpoint is the first to deliver a solution that enables organisations to find where their data is, understand what it is, and protect what both people and agents do with it.”
The new features are being rolled out gradually. AI exploit detection for email will be available from the fourth quarter of 2025. Proofpoint Data Security Complete is already accessible, with further updates coming over the next two quarters. Proofpoint Secure Agent Gateway and Satori Agents will enter phased release starting in 2026.
