Agent Commander reverses AI mistakes

Veeam Software has announced a new unified platform designed to help organisations detect AI-related risks, protect AI systems and recover from AI errors. Agent Commander aims to enable companies to address AI-driven threats proactively and scale AI agents securely.

Agent Commander represents the first integration following Veeam’s acquisition of Securiti AI. The system will be available in a future release of the Securiti Data Command Center. Early access is available upon request.

“AI happens at machine speed, which means organisations must understand what data is being used, by what agent, and how in real-time,” says Anand Eswaran, Veeam CEO. “If an error occurs, organisations not only need to understand what data was impacted, but they also need the ability to undo any damage rapidly.

“With Agent Commander, organisations know what data is powering AI, and it gives them the power to detect, protect, and, when necessary, undo AI actions with speed and precision. It represents the future of what’s expected from data security and data resilience, and it’s only possible with Veeam’s unified platform.”

According to Veeam, the most critical gap in AI infrastructure today is trust. As AI agents scale, data risk and AI risk have become the same problem. The company says an agent is only as trustworthy as the data it can see, access, and act on. Yet enterprise controls remain fragmented with separate systems for protection, security, governance, and recovery, and none built to provide unified visibility, granular control, or precision response at the speed and scale AI now demands.

Sensitive data is being fed into models and acted upon in ways no one approved or is tracking. As AI moves at machine speed, an AI agent can access and act on sensitive data. The company says traditional workflows that take hours to detect and days to remediate leave too much exposure at that velocity.

According to Veeam, closing this gap requires a new layer of AI infrastructure: a unified control plane that delivers contextual visibility, policy-level enforcement, and surgical recovery, converging data resilience, data security, and AI risk management into one operational system.

Agent Commander can provide organisations with visibility into their AI environments, detect hidden risks and Shadow AI, and apply controls to protect data as it moves through AI systems. The platform includes rollback capabilities designed to reverse actions taken by AI agents.

The system enables teams to detect and remediate threats, supporting safer and faster AI adoption. It is designed to provide the visibility and control required to scale AI securely across the business.

The Agent Commander core

At the centre of Agent Commander is Veeam’s Data Command Graph. This real-time relational intelligence engine maps live connections between data, identities, AI models, and autonomous agents across production and backup environments.

The platform is designed to identify areas of risk that other tools may not detect, including intersections between compromised identities, exposed data and autonomous agents. These overlapping risks can compound and evolve in real time across an organisation’s data and AI estate. Veeam positions Agent Commander as offering integrated visibility and control across data and AI environments, in contrast to standalone AI security or backup tools.

Veeam says the system delivers three capabilities the industry has never seen together:

• Detect AI Risk with Context – Identify shadow AI, sensitive data exposure, and risky agent behavior, with full visibility into downstream impact across systems and environments.
• Protect AI Pipelines Autonomously – Enforce granular, real-time controls across data, identities, and AI agents, independent of model providers, cloud platforms, and hybrid needs.
• Undo AI Mistakes with Precision – Surgically reverse unwanted AI actions using precise, context-aware recovery, restoring trusted data without reverting entire systems.

Rehan Jalil, Veeam president of products and technology, says: “As AI becomes operational infrastructure, enterprises can no longer treat data protection, data and AI security, privacy, and governance as separate disciplines. Controlling AI risk is effectively impossible in siloed environments or without deep contextual intelligence across data, permissions, and autonomous agents. Agent Commander unifies control across production and backup to detect toxic combinations, enforce granular policy, and precisely reverse AI-driven actions. This is the foundation required to operate AI safely at enterprise scale.”

Todd Thiemann, Omdia principal analyst for AIM and data security, says: “Acquisitions often raise the question of how combined strengths can create new customer value. Veeam’s announcement provides a clear roadmap for integrating capabilities and delivering enhanced solutions to organisations securing their data and AI estates. The focus on Agent Commander demonstrates Veeam’s commitment to provide both operational efficiency and comprehensive security for AI agents.”

* To request early access to Agent Commander, visit the website here.