Environment Agency trials AI tool to help sewage discharge permit assessors

The Environment Agency has begun a small pilot of the Regulatory Guidance Assistant (RGA), an AI tool built by Hitachi Solutions with in-house Defra digital teams to help water quality permitting officers find and apply regulatory guidance for sewage-discharge permits. The agency's transparency record states the RGA does not make regulatory decisions and that all AI-generated outputs require human review and remain subject to the permitting officer's judgement. The tool ingests and organises regulatory guidance through extract, transform and load (ETL) processes to reduce time spent manually searching fragmented documents, and is currently limited to a small group of staff focused on small sewage discharge issues.

The organisation, which is part of the Defra Group, has revealed a small pilot in which officials that work on water permitting are being given access to a new tool The Environment Agency is trialling the use of an artificial intelligence tool intended to help officers assessing applications for permits issued to water and sewerage companies. The organisation, which operates as an arm’s-length body sponsored by the Department for Environment, Food and Rural Affairs, has published a transparency record providing details on the operation of the Regulatory Guidance Assistant RGA algorithmic system. While it is not empowered or used to make decisions itself, the tool is designed to help officers assessing applications for sewage-discharge permits to understand the relevant regulations that should inform their decision. The transparency document describes the technology as “an AI‑enabled pilot tool developed within the Environment Agency’s Regulatory Services Programme to support officers by helping them find, navigate, and apply regulatory guidance more quickly and consistently”. During this pilot stage, the algorithm is being used by a small group test group of water quality permitting officers – whose role is to oversee the regime through which permits are issued to organisations engaged in “activities that have the potential to pollute” water, according to the agency’s careers website “This includes surface and groundwater discharges, groundwater spreading, dewatering activities and the necessary improvement schemes of water and sewerage companies,” the site adds. “We determine often complex applications, carrying out rigorous checks in process such as habitats risk assessments, modelling, consultations, pre-applications and more.” To help support the work of these teams, the new AI-powered system – which is currently only available to a small group of staff focused on “small sewage discharge” issues – is intended to “address long‑standing challenges with fragmented, inconsistently formatted guidance and inefficient retrieval processes”. The technology, which was built by supplier Hitachi Solutions working with in-house Defra digital teams, works by “ingesting and organising regulatory guidance, with materials that are extracted, transformed, and loaded ETL through ongoing… processes to ensure the tool works from up‑to‑date, structured source material”, according to the transparency record. The RGA then “uses AI to support more efficient and consistent discovery of policy and technical guidance, meaning officers can quickly retrieve authoritative guidance, which reduces time spent manually searching across multiple fragmented documents”. Related content - Environment Agency to launch online service to report water pollution https://www.publictechnology.net/2024/11/13/environment/environment-agency-to-launch-online-service-to-report-water-pollution/ - Environment Agency revamps digital tool for waste exporters https://www.publictechnology.net/articles/news/environment-agency-revamps-digital-tool-waste-exporters - Government flags up ‘Swimfo’ water-checking tool as sewage protests rumble on https://www.publictechnology.net/2022/08/26/society-and-welfare/government-flags-swimfo-water-checking-tool-sewage-protests-rumble/ The operational record stresses that the algorithm is focused on “providing assistance, not decisions”. “The RGA does not make regulatory decisions,” the document adds. “All AI‑generated outputs require human review and remain subject to regulatory judgement by the permitting officer.” The overall aim of the tool is to “reduce search friction, improve the speed and reliability of guidance discovery, and ensure greater consistency in how regulatory guidance is applied”. While initial pilots are ongoing, the Environment Agency “also aims to test the capabilities and limitations of AI in regulatory contexts – under strong governance and assurance”. Beyond the small-scale tests, other permitting officers at the agency are required to “locate and interpret regulatory guidance manually using a combination of SharePoint sites, guidance documents, intranet resources, search functions and personal knowledge.” “The process relies heavily on users understanding where guidance is stored, how documents relate to one another and which sources are most relevant to the specific permitting scenario being considered,” the record states. “This could require reviewing multiple documents and cross-referencing guidance from different locations before reaching a conclusion.” Before commencing with the AI trial, the agency considered the use of “traditional keyword and metadata-based enterprise search”, before proceeding with the RGA tool “because it combines natural-language querying with retrieval from approved guidance sources, enabling officers to locate and interpret relevant guidance more efficiently while maintaining human oversight and professional judgement”. The algorithm remains used solely in a closed-off production environment that is not connected to live systems elsewhere. But, in the longer-term, Defra hopes that the trial process will help deliver “reusable data and AI patterns for wider regulatory services… including patterns for data ingestion, ETL, guardrails, and governance which can support future services across the regulatory services programme and enable efficiency savings”.