Construction AI: leadership and vertical agents step out of the lab

Turner & Townsend has appointed Emma Gilthorpe as Chief Transformation Officer, with a mandate to drive global transformation across its operating platform and "embrace the opportunities of AI and technology". Gilthorpe was most recently Chief Executive of Royal Mail and previously spent nearly 15 years on the executive team at Heathrow, including as Chief Operations Officer through the pandemic.

This is the kind of hire that tells you AI has moved out of the digital team and into core operating-model strategy. Gilthorpe's brief is technology-led transformation across a large, distributed delivery business - exactly the problem most tier-one contractors and consultancies are still trying to organise around.

Reds10 Group has completed a strategic investment in Hull-based steel fabrication specialist ESL, framing the deal as part of a "next phase of strategic growth to exploit advancing technologies, while integrating AI at every level of the business" (Paul Ruddick, CEO). The company already operates fully automated lines for frame fabrication and panel assembly, with reported productivity gains close to 40 per cent in those facilities.

This is industrialised construction's playbook for AI in the open. Vertical integration tightens the data feedback loop, AI sits across design, manufacturing and delivery, and the commercial story is volume and predictability - not just productivity.

For Construction Safety Week 2026 (4-8 May), Turner Construction released SafeT Coach - an AI-powered safety assistant deployed inside ChatGPT - free of charge to other building professionals. Site teams can take a photo, ask a plain-language question, and receive guidance grounded in Turner's environmental, health and safety framework and OSHA standards. Turner reports more than 25,000 interactions with staff, trade partners and field teams since the initial pilot.

Two things stand out. First, a tier-one contractor is choosing to share its safety AI rather than treat it as a moat. Second, the deployment pattern is "agent inside an existing channel" - the same pattern UK ConTech entrants like BRCKS are using in WhatsApp.

On 5 May, Anthropic released ten ready-to-run Claude agent templates aimed at the most repeatable work in financial services: Pitch builder, Meeting preparer, Earnings reviewer, Model builder, Market researcher, Valuation reviewer, General ledger reconciler, Month-end closer, Statement auditor, and KYC screener. They ship as plugins inside Claude Cowork and Claude Code, plus a cookbook for Claude Managed Agents. At the same time, Claude is now embedded in Microsoft Excel, PowerPoint and Word as add-ins (Outlook to follow).

This is the productisation pattern construction will see next. Vertical agent templates, packaged inside the tools people already live in, with context that carries between Word, Excel and PowerPoint - that is the shape of how AI shows up at scale on cost consultancy, surveying and project controls work.

Cursor has opened a public beta of its TypeScript SDK, letting developers create, run and manage Cursor's coding agents from their own code, scripts, CI/CD pipelines and products - with sandboxed cloud VMs, subagents, hooks and token-based pricing. Agents launched through the SDK use the same harness, codebase indexing and semantic search as the desktop app.

ServiceTitan's 2026 Commercial Specialty Contractor Industry Report, drawing on a survey of more than 1,000 industry leaders, found that 38 per cent of contractors now see measurable business impact from AI - up from 17 per cent a year earlier. Cost estimation and budgeting (24 per cent) and bid management (22 per cent) lead the use cases, with reported automated estimating accuracy of 85-90 per cent against manual baselines.

The headline number is real movement in twelve months. The caveat is the gap between leaders and the rest: a separate Bluebeam survey found only 27 per cent of AEC professionals currently use AI, even though 94 per cent plan to increase usage in 2026.