The Environmental Consulting Playbook: AI Transformation for a Sector in Transition (Part 3 of 3)

In the first two parts of this series I argued that AI is restructuring consulting and that the firms who build defensible positions now will thrive. This piece is about one sector in particular: environmental and water consulting.

It’s the sector I know best. It’s also moving faster than most people realise.

What the Big Players Are Doing

Looking at what the large firms are doing reveals where the market is heading. Smaller firms don’t need to copy them to read the signal.

AECOM paid around $390 million last November for Consigli, a Norwegian AI startup whose tools claim to cut engineering time by up to 90%. Arcadis launched its own digital business unit (Arcadis Gen), and its automated solutions have reduced environmental report preparation time by more than 45%. RSK’s Witches Oak smart water project delivered £7.4 million in savings. WSP is expanding its Future Ready initiative and reportedly considering a multibillion-dollar merger with Jacobs.

The UK environmental consulting market is £4.1 billion and growing at 10.7% CAGR. RSK leads at £2.2 billion revenue and 13.2% market share. A majority of environmental consulting clients surveyed in early 2025 expect their providers to deliver at least a 15% improvement in project delivery times and cost savings through AI. The pressure is real.

Where AI Is Already Being Applied

Environmental Impact Assessments that used to take weeks of manual review now get processed in 2 to 4 hours. AI handles document synthesis, data merging, land-use classification from satellite imagery, and regulatory scanning with more consistency than a team of junior consultants.

Contaminated land shows where the technology hits its ceiling. AI can process thousands of historical studies in the time a human reads a handful. It can run jurisdictional comparisons (UK GAC standards against US EPA Regional Screening Levels) in a single query, handling unit conversions automatically. But it cannot tell the difference between old rust stains and fresh contamination. It recognises Japanese Knotweed shoots but misses Wildlife and Countryside Act implications. It classifies a water sheen as reflection rather than a hydrocarbon plume indicator.

These limitations are exactly where experienced environmental consultants add irreplaceable value. AI handles the data processing. The consultant provides the professional judgement.

The Risk Map

Not every service is equally exposed.

High risk (routine, standardisable, data-intensive): desk-based data review, standard Phase I Environmental Site Assessments, routine compliance monitoring, ESG data collection, carbon accounting, document review for transactions, routine GIS mapping, water quality data analysis.

Medium risk (AI does first pass, human judgement still essential): EIA drafting, contaminated land risk assessment, network modelling and asset management, air quality modelling.

Low risk (human judgement is the whole point): stakeholder engagement, expert witness work, strategic advisory, complex remediation design, ethical decision-making on environmental trade-offs, novel regulatory interpretation, field-based ecological surveys.

The revenue model under pressure is the one built on billable hours for high-risk services. When internal client teams deploy ESG or compliance software effectively, reliance on external consultants for routine tasks declines.

AMP8: The Biggest Opportunity in a Generation

If you work in the UK water sector, this is the moment.

Ofwat has approved £104 billion of combined investment for AMP8 (April 2025 to March 2030), doubling AMP7. That includes £12 billion to cut sewage spills from storm overflows by 45% from 2021 levels, £6 billion for nutrient pollution upgrades across around 1,000 sites, £3.3 billion for nature-based solutions, and £2 billion of development financing to unlock a projected £50 billion in additional sector investment.

Digital water solutions are growing at 18.4% annually in the US and Canada, with the European market expected to double to $27.2 billion by 2033. The Ofwat Innovation Fund has grown to £600 million, supporting 109 innovation initiatives across the sector. Documented results include Yorkshire Water cutting visible leaks by 57% at Hadfield, intelligent sensors reducing sewer overflow volumes by more than 80%, digital twins producing a 30% reduction in aeration energy use, and AI-supported leak detection delivering at least 50% cost improvements.

Companies spent most of 2025 recruiting and training. Intensive activity ramps up from 2026 onwards. The consulting opportunity is live now.

The Governance Gap

AI adoption in environmental regulation is accelerating faster than the policy to support it.

Regulators are already using it. In the US the EPA has expanded machine-learning models in chemical safety and pollution prevention. The FDA deployed “Elsa” for clinical protocol reviews. The DOE’s PermitAI tool builds on almost 30,000 documents to automate repetitive permitting tasks. In the UK, the Ofwat Innovation Fund is driving AI adoption across regulated water utilities, and AI systems automatically scan legislative databases for regulatory changes relevant to environmental projects.

Three issues remain unresolved: transparency and reproducibility of AI-generated insights; administrative record integration (no consistent protocols for documenting AI influence on screening decisions); and cross-agency consistency across fragmented standards. Regulated entities lack clarity on how AI outputs can be authenticated or contested.

For firms that understand environmental regulation and also understand AI limitations, this is a service line hiding in plain sight. AI governance advisory is ongoing work by nature: policies need reviewing, compliance needs monitoring, frameworks need updating. That’s recurring revenue.

Where to Start

The practical starting point is smaller than most strategy documents suggest. Pick one report type your team produces regularly. Time it end to end. Split the work into data gathering (where AI makes the biggest dent) and expert judgement (where your people add irreplaceable value). That pilot will tell you more than any strategy document about where the leverage sits in your specific business.

Beyond the pilot, the firms that will do best over the next three years will be the ones that:

• embed AI into delivery rather than treating it as a separate initiative
• identify and own a domain specialism deeper than any generalist can match
• build reusable frameworks and templates so every engagement produces IP
• develop AI governance as a service line alongside traditional project work
• position early for AMP8 if they operate anywhere near the water sector

The full white paper goes deeper: a service-by-service risk assessment, a ten-step action plan, the full AMP8 opportunity map, the startups to watch, and the governance gap in detail. Download it here.