Construction projects that incorporate artificial intelligence could see cost savings of up to 20% and schedule reductions of up to 25%, according to a study released Wednesday by Suffolk and the Massachusetts Institute of Technology. The findings quantify what early adopters are already reporting on active job sites - that AI-driven tools are cutting waste, accelerating timelines, and reshaping how projects are planned and managed.
The study, "Construction in the Age of AI," was conducted by Suffolk in partnership with the MIT Center for Real Estate and the MIT Media Lab City Science research group. Researchers analyzed a multifamily development completed in San Francisco between 2021 and 2024 and found that incorporating AI could have trimmed construction costs by 17% to 20% and reduced building time by 22% to 25%.
Those gains would boost a developer's profitability by 5 to 6 percentage points on unlevered internal rate of return and 1 to 2 percentage points for yield on cost, the report said. "That can be the difference between housing being built and housing not being built," the authors wrote.
Where the savings come from
The report identifies six areas where AI can cut costs: design automation, offsite manufacturing, permitting, scheduling, skilled labor and subcontracting, and supply chain and procurement. Seventy-five percent of construction projects currently exceed their budgets, and labor inefficiency wastes $30 billion to $40 billion annually in the U.S. alone, according to the study. McKinsey & Co. estimates that AI and generative AI could create $18 billion in value each year for U.S. homebuilders - roughly 10% of that sector's revenue.
Permitting stands out as a bottleneck ripe for automation. More than 20,000 independent permitting agencies operate across the U.S., creating fragmented application processes that drag out project timelines. AI can also coordinate long-established methods like offsite prefabrication and modular construction to reduce on-site congestion. Amid a shortage of skilled workers, AI-driven robotics can take on high-repetition tasks such as rebar assembly on factory floors, leaving human crews to focus on judgment-intensive work.
"AI presents a real opportunity to transform the way we build," John Fish, chairman and CEO of Suffolk, said in a statement. "Fully realizing this opportunity will require us to rethink how projects are planned, coordinated and delivered, enabling us to build with greater speed, efficiency and precision."
AI on active job sites
Several firms are already deploying AI outside of academic studies. Stockholm-based Skanska used robots from Boston startup Nextera Robotics during construction of Kaye, a 31-story, 324-unit apartment tower that opened last year in Seattle's Belltown neighborhood. The robots, called Didge, captured 360-degree images and video, feeding them into an AI model that flagged safety issues - water leaks, holes, exposed beams, misplaced fire extinguishers - and compared site conditions against preconstruction schematics and 3D digital blueprints.
Stewart Germain, director of innovation and sustainability for Skanska USA's commercial development arm, said the approach saved the team up to 40 hours per week that would otherwise have been spent walking the site and taking pictures manually.
Kirkland, Washington-based Cordillera Homes has used AI tools to streamline planning, design, and construction for projects including Central Peak Residences, a condominium development in downtown Kirkland. Michael Eney, the firm's chief financial officer, said the time savings in coordinating planning and design, creating schedules, and reviewing budgets are "a real game changer." He added, "Delivering homes more quickly saves money on every project."
Adoption varies widely across the industry. Larger contractors often make dedicated investments, while smaller firms gain access through construction software platforms such as Autodesk and Procore that are embedding AI tools into their products, said Erin Khan, a Los Angeles-based construction technology consultant and former Suffolk executive. "Even if smaller contractors don't have all the bells and whistles, they're still dipping their toe in the water and getting into it," she said.
Data sharing as the next hurdle
Khan, who was not involved in the Suffolk-MIT study, cautioned that AI is not a cure-all. Results depend on how well companies implement the technology and share data across project teams. "There's a ton of high potential, and if we can all get together and agree on how to share the data and what that entails, that would put boosters on AI for everyone," she said. "Contractors that understand the processes they are applying AI to are more likely to improve collaboration and communication across a job site."
James Scott, co-lead of the MIT Center for Real Estate, echoed the need for a broader evidence base. "The next step is to keep building the data foundation needed to understand where AI has the strongest impact, where the limits still are and how those findings can be translated into better decision-making across real projects," he said. "That kind of evidence base is essential if the industry wants to move from promising examples to durable, repeatable progress."
For professionals looking to build skills in this area, structured learning paths such as AI for Project Managers cover the scheduling, risk management, and workflow automation techniques that construction firms are beginning to apply. Broader AI for Real Estate & Construction resources address the preconstruction analysis and design coordination described in the Suffolk-MIT findings.
Why this matters for real estate and construction professionals
The 17% to 20% cost reduction figure is not a theoretical ceiling - it is derived from an actual completed project. For developers, that margin can determine whether a project pencils out. For project managers and site supervisors, the immediate takeaway is that firms already using AI for progress tracking, safety monitoring, and schedule optimization are reclaiming dozens of hours per week. The bottleneck is no longer the technology's capability but the industry's willingness to share data and integrate these tools across teams. The companies that solve the data-sharing question first will lock in a structural cost advantage that competitors cannot easily replicate.
Your membership also unlocks:
