Make Sets Out 'Glass-Box' Vision for Enterprise AI Adoption

What happened

At Make Waves 2026 in Prague, automation vendor Make outlined its vision for making increasingly autonomous business operations visible and controllable. The company framed this against the risks of "black box AI", setting out a "glass box" ambition for enterprise AI adoption, according to coverage by Diginomica.

The presentation positions transparency and governance as the conditions under which businesses can take on more automated work. The reporting centres on Make's argument rather than on a new product release, so the announcement is best read as a statement of direction from the vendor.

Why it matters

As AI moves from assisting people to acting on their behalf inside workflows, the question for leaders shifts from whether to automate to how they can see and steer what automated systems are doing. A glass-box approach implies that operations teams, compliance functions and frontline staff need to understand why an automated process took a given action, not just that it completed.

For digital transformation programmes, this places observability and control alongside speed and cost as design requirements. Organisations that cannot explain or override automated decisions may find it harder to scale them, particularly where customers or employees depend on the outcome.

The Renascence take

Most discussion of autonomous AI focuses on capability, but the experience question is about legibility. People tolerate an automated decision far more readily when they can tell what it did and why, and that principle holds whether the person is a customer, a claims handler or an operations manager.

The real test of "glass box" AI is not whether a dashboard exists but whether a frustrated customer or an overruled employee can get a clear answer in plain language. Treat visibility as a service-design requirement from the first build, not a reporting feature added later. Map every point where an automated action changes an outcome someone cares about, and decide in advance who can intervene and how. Operators that do this will earn more autonomy over time; those that don't will find that every new automation quietly erodes trust.