David Taylor
Chief Commercial Officer
Feb 23, 2026
Artificial Intelligence is the new cure-all.
Missed forecasts? Add AI. Cost overruns? Add AI. Disparate systems? Layer AI on top. That’s the narrative.
Here’s the reality: AI will not fix your project controls. In fact, if your foundation is weak, it will amplify the problem — faster and at scale.
Most capital programs today still struggle with inconsistent WBS structures, misaligned cost codes, manual data reconciliation, and reporting cycles that rely on spreadsheets stitched together the night before the review. Data governance is often implied, not enforced. Ownership is blurred. Definitions vary by contractor.
Now imagine introducing machine learning and AI into that environment. What exactly is it learning? AI models depend on reliable and consistently governed data. If your baseline shifts without discipline, if forecasts are adjusted to “manage optics,” if actuals lag by weeks, the algorithm doesn’t correct those behaviors. It absorbs them.
Garbage in... faster, more confident garbage out.
There’s another issue few talk about: organizational behavior. AI can detect variance patterns. It can flag anomalies. It can predict potential cost or schedule drift based on historical trends. It cannot fix optimism bias. It cannot override risk aversion. It cannot resolve political forecasting inside large programs.
If project leaders are hesitant to surface bad news early, AI won’t suddenly create a culture of transparency. It may actually create a false sense of objectivity — “the system says we’re fine” — when the underlying inputs are flawed.
That’s not innovation. That’s automation of ambiguity. This doesn’t mean AI has no place in project controls. It absolutely does. But it requires maturity first.
Before AI and predictive analytics, you need:
Standardized taxonomies across cost, schedule, and risk
Clear system ownership and governance
Integrated data flows, not batch exports
Defined accountability for forecast accuracy
Unified historical datasets
In other words, infrastructure.
Too many organizations want advanced intelligence layered on top of fragmented ecosystems. That’s like installing a high-performance engine in a vehicle with a misaligned chassis. It may run — briefly — but it won’t run well.
There’s also a strategic risk. AI-generated insights carry authority. Executives trust them. Boards lean into them. If those insights are built on unstable foundations, decision-makers become more confident — not more cautious. That’s dangerous.
The organizations that will benefit most from AI in project controls are not the ones chasing headlines. They’re the ones quietly standardizing, integrating, and governing their data environments first. They treat AI as an accelerator — not a crutch.
Project controls has always been about discipline. Measurement. Accountability.
Transparency. Technology can enhance that discipline. It cannot replace it. So before asking, “How can AI improve our forecasting?” ask a harder question: Is our data environment mature enough to deserve AI? If the answer is no, the path forward isn’t another tool. It’s getting the fundamentals right — and building the kind of project intelligence ecosystem that AI can actually strengthen, instead of expose.