AEC Magazine Q&A: Managing AI Coding Risks in AEC
AI coding · Enterprise AI · BIM collaboration
In an interview with AEC Magazine, AMC Bridge CEO Igor Tsinman discussed the surge of AI-generated code in AEC firms, addressing governance needs, maintenance costs, and a blended development model.
AEC Magazine interviewed Igor Tsinman, CEO of AMC Bridge, on the rapid rise of AI-assisted software creation across AEC practices. Tsinman noted that AI tools allow non-programmers to quickly create functional prototypes to address specific business friction. However, moving tools into production requires testing, security audits, and lifecycle maintenance. AMC Bridge's internal research indicates that while AI can cut routine coding time by 70 to 80 percent, efficiency gains remain much smaller for complex 3D geometry and specialized engineering systems.
Key takeaways
- AMC Bridge's internal research suggests AI reduces routine coding time by 70 to 80 percent, but offers less improvement for complex 3D geometry and engineering math.
- Tsinman advocates a blended approach: purchasing core commercial platforms for 80 percent of workflow needs while developing targeted tools for the remaining custom gaps.
- Unregulated internal tools risk altering live models; recommended governance includes sandbox testing, audits, release gates, and freezing unverified tools with write access.
- Prototypes that support recurring workflows require designated owners, corporate repository storage, credential security, and defined maintenance plans.
Source：AEC Magazine Read the original ↗
Prepared, checked and automatically published by AI; not individually reviewed by a human.