The short version: The American Society of Civil Engineers released its AI Roadmap for Application in Civil Engineering, called AI RACE, in October 2026. It is the profession’s first broad plan for how civil engineers should use AI, and its core message is simple: AI can help with almost every stage of a project, but the licensed engineer keeps responsible charge and the final say.
What happened
ASCE pulled together subject matter experts over 2026, held an AI Roadmap Summit on August 31 and September 1, finalized the roadmap at a two-day workshop in September, and won support from its Board of Direction on October 17. The advisory committee was chaired by Yinhai Wang, a transportation engineering professor at the University of Washington.
ASCE says civil engineers have been adding AI tools on their own, firm by firm, without shared standards or rules. AI RACE is meant to give the whole profession one direction.
What is in the roadmap
- A clear structure. A foundational layer plus 11 pillars grouped into four domains, covering planning, design, construction, operations and maintenance.
- A maturity path. Short, medium and long-term stages that move from assistive AI toward more autonomous AI over time.
- Human red lines. Five plainly stated limits on what AI must not do, plus a proposed matrix showing where the line between AI and human work sits.
- Responsible charge throughout. Every use of AI is judged by whether it strengthens or weakens the engineer’s responsibility for the work. Accountability to the public always rests with identifiable people.
- Ethics and licensure. One pillar covers ethics, governance, licensure and public trust, including open questions on liability, intellectual property and data ownership.
What ASCE plans to do next
- Set standards and guidance on intellectual property, data rights and data sovereignty in engineering practice.
- Write standards for AI-assisted project controls, digital twins and autonomous data capture.
- Publish machine-readable versions of key ASCE standards.
- Develop practice standards for siting, designing, powering, cooling and securing AI infrastructure such as data centers.
- Offer AI training through continuing education credits, certificates and webinars, with support for small firms and public agencies.
Why it matters now
AI is already in civil work. Generative grading tools, design assistants in Civil 3D, AI-driven inspection and digital twins of water and transportation systems are in use at large firms. Two-thirds of the finalists in Bentley’s 2026 infrastructure awards reported using AI. What has been missing is agreement on who is responsible when an AI-assisted design goes wrong. AI RACE answers that the same way the profession always has: the engineer who stamps it.
The data center piece is worth noting too. ASCE now treats the facilities that run AI as civil infrastructure in their own right, with their own power, water and siting questions. That is a lot of future civil, structural and site work.
Who should pay attention
Civil and structural engineering firms, DOTs, water utilities and public works departments, and land development teams. Expect clients and insurers to start asking how firms use AI, and to point to ASCE’s framework when they do.
What it means for hiring
ASCE has made it clear that AI does not change who is responsible. That keeps licensed engineers at the center of every project.
- PE-licensed civil engineers remain the hard-to-fill hire. Responsible charge cannot be handed to software.
- AI literacy is becoming a real hiring factor. Engineers who can use these tools and explain their limits will move faster.
- Firms will need people who can set internal AI policy and QA steps. Senior engineers with that judgment are worth recruiting for.
Hiring civil talent? See our civil engineering recruiters page, or read is structural engineering a good career.
Interview questions for civil engineers
- Which AI tools have you used on real projects, and where did you not trust the output?
- How would you check an AI-generated grading plan or calculation before sealing it?
- What should a firm’s AI policy say about client data?
- What does responsible charge mean to you on a project where software did much of the drafting?
Frequently asked questions
What is ASCE AI RACE?
AI RACE is the American Society of Civil Engineers’ AI Roadmap for Application in Civil Engineering, a framework for using AI across the infrastructure lifecycle while keeping engineers accountable for the work.
Does AI RACE allow AI to design infrastructure?
It supports AI-assisted design, but sets firm limits and keeps responsible charge with the licensed engineer, who has the final say.
Will ASCE offer AI training?
Yes. The roadmap says ASCE plans continuing education credits, certificates and webinars on AI, plus help for smaller firms and public agencies.
More stories like this are on our AI for Engineers hub.
