Kathy Gibson reports from the Veeam ProPartner Summit 2026 – Artificial intelligence (AI) has shifted enterprises’ relationship with their data, which could expose that data to new threats.
To plug this emerging gap in enterprise resilience, Veeam has added a new DataAI platform to ensure trust in the AI infrastructure.
Veeam’s lifecycle has straddled three eras since its inception in 2005, and the company continues to innovate and evolve to ensure it stays relevant.
This is the word from Mena Migally, regional vice-president: EMEA East at Veeam, who explains that the company was created to provide fast virtual machine (VM) recovery. “We made sure that, whatever happened, we were able to bounce back,” he says. “We made sure that, wherever the data was, protection followed.”
The second era was cyber resilience, acknowledging that breaches were going to happen and ensuring it could provide a comprehensive end-to-end response.
The latest era – agentic AI – ushers in a world of autonomous agents, operating at machine speed.
“In this era, we are looking at three triggers colliding all at once,” Migally says.
Trigger one is that data gravity flips. In the old way of working, each app was siloed; but agentic AI unlocks all of them, reaching into every system of record.
“This also speaks to scaling,” Migally says. “Data that was previously dormant is now the fuel for AI, which runs on 90% of an enterprise’s data compared to the 10% of structured data that was previously used.”
This reach means agentic AI can also be weaponised. “And a failure today isn’t simply a breach – it is decisions happening at machine speed,” Migally points out. “Before you get an alert, or see an anomaly on your dashboard, it has happened.”
Trigger two is that fact that the infrastructure to deploy AI is going parabolic, with billions of dollars being spent today by hyperscalers – and trillions set to come in the years ahead.
“The autonomous workforce isn’t coming – it is here,” Migally stresses. “Today, there are 82 agents for every one human in the average enterprise, with 54% of enterprises running AI agents in production today.”
We need to remember that these agents don’t operate at human pace, or keep human hours. “They operate on different rules: frequency and volume of data. Speed collapses the window for attack response, and the blast radius is no longer limited. “It doesn’t expand, it explodes because 97% of agents carry excessive privileges,” Migally says. “This isn’t theoretical, it is happening today.”
With all of this change, the role of remediation has had to change as well.
“When things go wrong, you cannot just roll back the entire enterprise – you need to contain and recover in a very precise manner,” Migally points out.
“These are not questions that we had to consider a year ago. And they are not being asked by just IT people – but in governments, boardrooms and every part of society.”
While the infrastructure to deploy AI has gone parabolic, infrastructure of trust hasn’t.
“Because zero trust was built for humans, not for AI,” Migally explains. “The control point now has to be the data itself, not the people.”
The AI stack today consists of compute, data, models, agent orchestration, and applications and vertical agents.
“But one layer is missing, the most consequential layer in AI: trust in the data.”
For this trust to be real, enterprises have to understand their data at a granular level; they need complete context that travels with the data; this context and identity-aware runtime enforcement must happen long before any action is taken; and they must be able to perform precision remediation – “when something goes run you must be able to isolate it and remediate it”.
Veeam recently made its biggest investment to date, with its acquisition of SecuritiAI.
“With this, we have created the data and AI trust category,” Migally says. “It sits in the stack above data and analytics, ensuring that data is safe at scale, with data access control, identity and AI coming together in one platform.”
The Veeam DataAI Command Platform is brought to life through the DataAI command graph, enabling resilience, security, governance, compliance and privacy in one combined platform.
“Veeam is building the infrastructure to provide DataAI trust,” Migally concludes.