The short version

Shadow IT used to mean someone sneaking in a file‑sharing app or a rogue scheduling tool. Now it’s shadow AI: project engineers pasting contract language into unapproved chatbots, site supervisors feeding drone imagery into free image analyzers, estimators testing cost models on public tools.

The Forbes piece on curbing a culture of shadow AI isn’t written for construction, but the subtext lands squarely on the jobsite: people are already using AI tools, whether or not your company has a policy—and the risk isn’t theoretical.

The real danger isn’t that people experiment with AI, but that they do it without guardrails, oversight or any shared understanding of what’s at stake.

The article argues that leaders who simply ban tools like ChatGPT or other generative platforms don’t stop usage; they just drive it underground. Instead, it calls for a deliberate culture shift: clear guidelines, open dialogue, and sanctioned paths to experiment with automation.

Translate that into construction technology and you get a blunt reality: if you don’t define how AI in construction can be used, your teams will define it for you—on live projects.

Why this matters on real projects

The Forbes argument revolves around three ideas that map neatly onto construction:

1. **People are using AI because it actually helps.** Office staff lean on AI tools for emails, RFIs, method statements, and change order narratives. Field teams are starting to ask generative models for safety brief templates or sequencing suggestions. Shadow AI doesn’t appear out of nowhere; it appears where the pressures are highest—compressed schedules, thin margins, and overworked coordinators.

2. **Unmanaged AI usage can leak data and amplify errors.** The piece highlights the risk of employees feeding sensitive information into public models. On a project, that could mean: - Contract clauses copied into a chatbot to “simplify the legalese.” - Bid numbers or proprietary production rates dropped into an online estimator. - Site photos with identifiable workers uploaded to an unvetted image tool.

Once that data leaves your perimeter, you don’t control where it’s stored or how it’s used. For a contractor, that’s not just an IT problem—it’s a commercial and reputational risk.

3. **Culture beats policy every time.** The Forbes piece stresses that fear‑driven bans don’t work; people find workarounds. In construction, that’s doubly true. Project teams are paid to “make it happen,” and they’ll quietly use whatever automation shaves hours off paperwork.

Instead, the article advocates: - **Explicit guidelines** on what data can and cannot be shared with AI tools. - **Approved platforms** that are configured for enterprise use. - **Training** so people understand both the upside and the limits of AI.

For construction leaders, that translates into practical moves:

• Draft a short, plain‑language AI use policy specific to project work: drawings, models, photos, contracts, and safety data.
• Identify a small set of vetted AI tools—maybe a contract‑review assistant or a summarization bot for meeting minutes—and make them easy to access.
• Encourage teams to log their AI experiments: what they tried, what worked, what failed. That turns shadow AI into a source of innovation instead of a hidden liability.

The tension the Forbes article surfaces is this: AI in construction is both a productivity boost and a new class of risk. The companies that win will be the ones that treat it as a managed part of their construction technology stack, not a forbidden side hustle.

What to watch next

• **Formal AI playbooks:** Expect more contractors to publish internal AI guidelines that spell out approved use cases, from drafting correspondence to assisting with schedule narratives.
• **Vendor promises vs. reality:** Software vendors will lean harder on "AI‑powered" marketing. Scrutinize how they handle your project data, model training, and audit trails.
• **Data governance on the jobsite:** Conversations about safety and quality will start to include where AI tools are allowed to touch photos, RFIs, and as‑built models.
• **Role of the AI champion:** New internal roles will emerge—operations or VDC leaders tasked with turning scattered AI experiments into structured, safe automation.
• **Union and workforce reactions:** As AI in construction grows, expect sharper questions from labor about surveillance, deskilling, and who benefits from the productivity gains.

Field note from the editor

When I talk to supers and project engineers, they rarely ask *whether* they should use AI—they ask if they’ll get in trouble for admitting they already do. That’s the gap the Forbes piece is really pointing at.

Construction has lived through this before with Excel macros, drone flights, even WhatsApp groups for coordination. We ban them, then quietly depend on them.

Shadow AI is the same pattern, just with higher stakes. If you’re leading a project or a business unit, the question isn’t, “How do I stop people from using AI?” It’s, “How fast can I give them a safe, transparent way to do it?”

That’s not about chasing the latest gadget. It’s about owning the automation that’s already shaping how your jobs get built.