Why context is missing for trustworthy AI to work

According to M-Files CEO Jay Bhatt, the pace and complexity of work is changing the way organizations operate.

In a recent interview, he said work now spans more systems, teams and external partners, and “rarely follows a clean, linear path.” At the same time, leaders are under pressure to move faster, maintain compliance, and make decisions in real time, even as many operating models are designed for simpler times.

This mismatch is one reason why companies are pouring money into automation and AI, but performance feels stagnant. Butt argues that technology does not automatically eliminate complexity, but in some cases makes the underlying problems more visible. When information is fragmented or difficult to trust, teams waste time searching for documents, reconciling versions, and validating decisions. This resistance creates operational friction that multiplies between departments and limits the benefits of even the most modern tools.

Bhatt describes documents as the “DNA” of modern business because they contain the essence of how an organization operates. he said: “Contracts, policies, project files, and client records are not just byproducts of work. teeth the work. These support operations, compliance, and accountability. When documentation is not properly managed, everything built on top of it becomes slower and more vulnerable. ”

Treating documents as static files starts to cause problems, he says, because teams lose context. People cannot easily understand why decisions are made, who owns the action, or how the work is connected to broader processes.

This leads to delays and rework, but it also creates risk because there is a loss of trust in both the information and the systems that manage it. Eventually, staff will default to workarounds, manual checks, and knowledge that remains in people’s heads.

Although AI has become a major initiative in enterprise transformation, Bhatt believes the impact will be elusive because AI is only as good as the information underneath it. In many organizations, information is scattered across systems, inconsistently managed, and unevenly governed. Without a strong foundation, AI can produce output that is difficult to trust or explain, expected acceleration turns into doubt, and efforts can stall after initial experiments.

M-Files promotes a “context-first” approach, which Bhatt says starts with how information is used, not where it is stored. The idea is to automatically connect documents to supporting work (clients, projects, obligations, decisions), reducing the time your team spends searching, reconciling, and second-guessing.

“For teams, this means less time spent searching, adjusting, and second-guessing. For leaders, it provides an enterprise knowledge graph that increases visibility and trust.”

“You can see how work is progressing, where risks are elevated, and whether policies are being followed without relying on manual updates or organizational memory. And importantly, this has to be done in tools that people already use every day (like Microsoft 365), otherwise adoption becomes a bottleneck. Doing so can reduce operational friction at scale.”

Looking forward, Bhatt argues that a key shift in thinking is to treat operational friction as a strategic issue and the situation as an antidote. Support speed, compliance, and AI readiness at the same time by capturing context across your organization.

Read the full interview here.