Is AI a Double-Edged Sword for Engineering and Construction Stocks?

As design, engineering, and construction firms survey the potential impact of changes brought by artificial intelligence, their share prices have reacted in various ways.

Fears around AI disruption have dragged down the stocks of firms focused on design, which face elevated uncertainty owing to technological disruption risk and their billing structure. The market worries that an AI “efficiency trap”—where increased efficiency owing to AI-related time savings cannibalizes revenue by reducing billable hours—could disrupt the billable-hours model that many design firms use.

Yet, the potential upside from the data center buildout has boosted the shares of engineering, procurement, and construction firms. Because these firms handle a project’s full lifecycle, hourly billing isn’t an issue. These firms are also viewed as less risk-exposed because physical construction work is less vulnerable to AI disruption.

Which Factors Give Some Firms a Boost

Over the next few years, we expect billing structure will be a key factor in determining which companies benefit from AI efficiencies, as fixed-price contracts offer protection against billable hour compression. But in the long run, we expect 100% of AI-driven savings will be passed through to clients because engineering firms lack economic moats.

In addition to billing arrangements, several other factors will separate AI winners from losers:

• The nature of the work, with engineering more risky than physical construction.
• Contract mix, with fixed-price contracts being less risky than reimbursable ones.
• End-market exposure, with mission-critical end markets such as nuclear less vulnerable to AI disruption.
• The use of proprietary AI and datasets.

The Market Is Overestimating AI’s Upsides and Downsides for These Firms

Most of the engineering firms we cover have relatively limited exposure to data centers, still in the low-single digits, but the business is growing rapidly, and we expect the end market to become an important revenue growth driver over the next decade.

We expect strong demand and backlog growth in the near term, but we see current tailwinds as cyclical rather than permanent. We think the market is overly excited about the potential upside from the data center buildout for these names, extrapolating current elevated revenue growth rates well into the next decade. Furthermore, we expect additional contractor capacity and growing competition to gradually pressure pricing power and constrain long-term margin expansion.

We acknowledge AI’s potential for disruption and believe that engineering firms will need to adjust their models to navigate this environment. That said, we don’t see the risk as existential because engineering firms offer valuable expertise and help ensure regulatory and safety compliance, especially in regulated end markets. We think engineering firms will likely try to shift their contract mix toward more fixed-price contracts to reduce the risk of compression in billable hours because of AI.

The rise of AI is not the first technological advancement that has threatened to disrupt the billable-hours revenue model. Despite fears of reduced billable hours, technological improvements have expanded project scopes, raised safety standards, and increased the number of projects for engineers. In addition, the number of engineers has continued to increase steadily despite major technological innovations such as the introduction of computer-aided design and building information modeling software.

In the near to medium term, we believe AI tools will enhance engineering firms’ productivity and efficiency, thereby boosting margins. We are still in the early stages of AI adoption in engineering, and there is little visibility into how payment structures will evolve. If AI makes engineers more efficient and reduces billable hours, it can create a revenue headwind unless engineering firms pivot to charging clients based on outcomes rather than hours spent. AI-driven productivity gains could follow a similar path to cost savings from offshoring, with engineering firms and their clients sharing the cost savings.

Which Engineering Firms Are Poised to Weather the Storm?

We see specialty construction players like Quanta PWR and MasTec MTZ as best positioned to navigate AI issues. They are both the least vulnerable to disruption and have the greatest upside from participating in the data center buildout.

On the other hand, we see engineering firms such as Aecom ACM, AtkinsRéalis SNCAF, and Jacobs J as the most vulnerable. The nature of their work makes it theoretically more likely to be automated by AI, and they have relatively less exposure to data centers.

We think Fluor FLR sits in the middle: Its traditional EPC work mostly shields it from AI disruption (except for front-end engineering), but its AI investment upside is more limited than that of specialty construction firms.

That said, our top pick in the space is Aecom.

As a result of concerns around its AI strategy, Aecom is trading at a discount both to its own historical enterprise value/EBITDA multiple, as well as at a historically large discount to its peer Jacobs. Both companies have similar business models, and Jacobs has relatively higher exposure to reimbursable contracts, which could face the risk of billable-hour compression owing to AI-driven efficiencies. We don’t see a nearly 40% valuation discount to Jacobs as warranted.