Bentley Systems is accelerating its push into artificial intelligence, embedding AI capabilities across its infrastructure software portfolio to help engineers design, model and deliver projects more efficiently while retaining full control over decisions and outcomes.
At its recent Year in Infrastructure (YII) event in Amsterdam, the company outlined how AI is becoming a core part of its civil design, subsurface modelling, construction planning and asset lifecycle tools. Rather than treating AI as an add-on, Bentley is positioning it as a foundational shift in how engineering data, models and organisational knowledge are connected.
According to Bentley, AI will act as a practical companion for engineers—automating repetitive workflows, exploring thousands of design options, interrogating models using natural language and offering recommendations grounded in engineering logic, standards and real-world constraints. Crucially, the company emphasised that engineers remain firmly “in the loop”, with full visibility and authority over AI-generated outcomes.
Speaking at YII, Bentley Systems CEO Nicholas Cumins highlighted projects where AI-supported workflows have reduced schedules by as much as 80% and enabled levels of design optioneering previously considered impractical. At the same time, he acknowledged that AI adoption remains in an experimental phase for many firms.
“Not every project that adopts AI sees transformative results yet,” Cumins said. “Many teams are experimenting, learning and iterating—but that’s exactly how progress happens.”
Bentley’s AI approach centres on providing strong context awareness. AI tools are designed to understand engineering intent, performance goals and constraints such as budget, safety, carbon impact and regulatory requirements. They also incorporate real-world environmental data, drawing on Bentley’s reality modelling and subsurface analysis technologies, including iTwin Capture and Seequent.
A key pillar of the strategy is what Bentley calls an organisation’s “collective memory”—using decades of project data, standards and documentation to support AI-driven assistance. Bentley reiterated that customer data remains fully owned by customers, who decide if and how it is used to train AI tools.
This data strategy is underpinned by the Bentley Infrastructure Cloud, which unifies models, documents, sensor data and operational information in open formats. Bentley believes this lifecycle-wide data foundation is essential for AI to deliver reliable design, construction and operational insights.
At the centre of the AI roadmap is Bentley’s new generation of “Plus” applications, including OpenSite+, OpenUtilities Substation+ and Synchro+. These iTwin-native applications are built specifically for AI-enabled workflows and write data directly to intelligent digital models. OpenSite+, for example, automates drainage design and earthworks optimisation, allowing engineers to evaluate thousands of grading scenarios in seconds.
Bentley also showcased Bentley Copilot, a context-aware AI assistant that is being rolled out across both new and established products such as OpenRoads Designer and OpenRail Designer. Bentley Copilot enables users to query documentation, interrogate models, execute commands and make complex design changes using natural language, while maintaining a full audit trail of decisions and modifications.
The company stressed that AI recommendations are always constrained by engineering principles and design codes. “Left unchecked, a purely data-driven AI might propose solutions that appear optimal in simulation, but are unsafe in the real world,” Cumins warned.
Looking ahead, Bentley is inviting customers to help shape its AI roadmap through the Infrastructure AI Co-Innovation Initiative, while also exploring new commercial models that reflect the balance between human expertise and AI-driven automation.
With its expanding AI portfolio, Bentley Systems is aiming to redefine how infrastructure is planned, delivered and operated—combining automation and intelligence with engineering judgement, accountability and trust.
