Veeam says that it's adding AI agent trust to its existing data trust capabilities
The company says that it knows about data trust from its first two eras; backup and recovery from 1980 to 2017, and cyber-resilience from 2017 to 2026. That was with backup data. But now data trust in the AI era has to include production data, and AI agents, digital employees have to be trusted, to operate within an environment that makes them trustworthy.
We are at the dawn of the agentic AI era and there is no agent trust infrastructure. Veeam believes it can build one by using its acquired Securiti technology and integrating that with its existing protection and cyber-resilience software.
This is being extended from structured to massive amounts of unstructured data while AI infrastructure is being installed and deployed at an accelerating rate. Data trust has been about human access but now AI agents, digital employees, are operating at machine speed and the infrastructure to assess and verify their access to data largely does not exist - yet they vastly increase data access rates and can do enormous damage in seconds.
We need, Veeam says, a zero trust agent SW capability that provides granular, multi-domain context across data, identity and AI. This should be part of the metadata that accompanies data and enables real-time, agent identity-based, data access guardrails. Agent identity access management (AIAM) can find gaps and mishaps and fix them.
There needs to be, it says, a Data and AI Command Platform. Within this the DataAI Command graph enables security, governance, compliance and privacy as part of AI data resilience. We can, Veeam says, find DataAI risks and eliminate them.
This has to be done by AI agents as they can run continually, monitoring privacy, compliance and access consent, with evidence that can be provided to audit operations. When you have thousands of AI agents you cannot rely on human monitoring of their activities. The scale of agent activity will make that impossible.
For example, privacy operations can be automated using agents who are given regulatory context. Veeam’s coming v13.1 Data Platform will enhance its foundation data protection and cyber-resilience capabilities. It will Integrate AI to improve product capabilities a.eg; AI-enabled admin assistance, data security and data resilience operations. There will be an MCP server so your preferred chatbot can be used to “talk” to Veeam backup data and provide AI-powered threat response, audit intelligence, instant answers to backup queries and an overall AI experience that connects security and protection.
Read more about v13.1 capabilities here.
Bootnote
Veeam bought Securiti for $1.7 billion in October last year, gaining knowledge graph-based data security posture management (DSPM) that covers both primary and secondary data. Veeam's Securiti scans a production data set, such as a customer's entire AWS data estate, and maps it using graph technology. Then organizations can locate sensitive data, run a ROT (redundant, obsolete, and trivial data) analysis, make a breach impact analysis, check for compliance and related controls, inspect access controls and AI governance, and look at the data's protection status.
At the time Veeam CEO Anand Eswaran said: “We’ve entered a new era for data. It’s no longer about just protecting data from cyber threats and unforeseen disasters; it’s also about identifying all your data, ensuring it’s governed and trusted to power AI transparently… We bring those capabilities together in a single solution to help customers understand, secure, recover, and rollback, and unleash their data to drive new business value.”
Securiti CEO Rehan Jalil said: “In the AI era, security will need to be data-centric, as there would be no AI security without data security.”
He is now Veeam’s President for Products and Technology.