FOR GB ENERGY SUPPLIERS

AI assurance, mapped to Ofgem.

GB energy retailers are deploying AI into customer journeys, vulnerability detection, debt management, dynamic tariffs and grid forecasting. Each use case creates a regulator-facing assurance question. Probari turns those questions into a live evidence trail — one tenant, every use case, every framework, one source of truth.

Book a 30-minute pilot call See the framework stack

FOR HEADS OF AI · CHIEF RISK OFFICERS · DPOs · COMPLIANCE

THE PROBLEM

Every AI use case is now a regulatory question.

A GB energy retailer in 2026 is carrying a stack of obligations that did not exist five years ago. Ofgem's Consumer Standards regime sets explicit conduct obligations on automated decisioning. The EU AI Act applies to systems that touch EU citizens or markets. UK GDPR covers any model trained on customer data. ISO 42001 is becoming a board expectation.

Most retailers have an AI register on a spreadsheet, an AI Impact Assessment (AIIA) template in a Word document, and a control matrix nobody can find. Probari brings all three onto one platform with a sign-off chain, a citation trail and a live signals feed. The work happens once and stays current.

WHERE AI IS LANDING

Six common use cases. One assurance pattern.

01

CHURN PREDICTION

Models that predict which customers will switch. Touches retention pricing, vulnerability, fairness and the Consumer Duty crosswalk where applicable.

02

DEBT & ARREARS

Affordability scoring and repayment-plan recommendation. High vulnerability sensitivity, direct Consumer Standards exposure and tight lawful-basis requirements.

03

DYNAMIC TARIFFS

Time-of-use and personalised tariff recommendations. Touches transparency obligations, smart-meter data lawful basis and fair-pricing principles.

04

DEMAND FORECASTING

Models feeding wholesale procurement and balancing positions. NESO Grid Code touchpoints, operational resilience under NIS2.

05

VULNERABILITY DETECTION

AI flagging customers in financial or health-related vulnerability. Highest regulator-attention category; demands airtight explainability and review.

06

CHATBOTS & AGENTS

Generative customer-service and field-engineer copilots. Risks around inaccurate output, prompt logs and vendor-model passthrough — all assessable in Probari.

THE FRAMEWORK STACK

Eight in-scope frameworks. One control library.

01

UK GDPR & DPA 2018

Lawful basis, data minimisation, special-category data, automated decisioning rights, retention.

02

EU AI ACT

Risk classification, prohibited practices, high-risk obligations, GPAI passthrough — applied where in-scope.

03

ISO 42001

AI management system controls — policy, governance, lifecycle, supplier management, evaluation.

04

OFGEM CONSUMER STANDARDS

Domestic and non-domestic supply standards covering fair treatment, vulnerability, complaints and clarity.

05

NESO GRID CODE

Operational obligations where AI feeds balancing, forecasting or system-operator interactions.

06

NIS2

Operational resilience for essential services. Applies to AI components in critical operational paths.

07

EQUALITY ACT 2010

Non-discrimination obligations relevant to scoring, pricing and decisioning models.

08

CONSUMER DUTY CROSSWALK

Where retail energy products carry Consumer Duty-adjacent obligations, the crosswalk is mapped.

HOW PROBARI DELIVERS

The four-pillar method, configured for GB Energy.

The GB Energy tenant ships pre-configured with the eight-framework stack, the Ofgem-aligned AI Impact Assessment (AIIA) template and a control library that maps every question to evidence. The Northbrae Energy reference deployment is live in the demo environment.

DESIGN

Use case intake

Every new AI use case enters through the Probari intake wizard. Risk tier is automatically graded against the EU AI Act schedule and Ofgem-relevant categories at the point of capture.

  • Purpose, lawful basis, data classes, decision affected
  • Automatic risk grading (aligned to EU AI Act)
  • Stakeholder map and owner sign-on

ASSESS

AI Impact Assessment

Forty-seven AIIA questions across nine sections, drafted with Perplexity Sonar assistance — the only AI provider Probari uses, chosen because every answer comes with the source — reviewed by the use-case owner and signed off by the second line. Every answer carries a citation, an author and a timestamp.

  • AIIA template aligned to Ofgem expectations
  • Perplexity Sonar-assisted drafting
  • Citation chain to source documents

ASSURE

Control library & evidence pack

Controls mapped across all eight frameworks. Evidence pack exports as a single bundle for board, regulator query, internal audit or third-party assurance.

  • Eight-framework crosswalk
  • Evidence pack export
  • Audit-ready trail per use case

OPERATE

Live signals & reassessment

Operational signals tracked against the assurance baseline. Drift in input distribution, fairness metrics or customer-outcome indicators opens a reassessment ticket — not a surprise audit.

  • Signals feed
  • Drift detection
  • Reassessment queue

WHO IT IS FOR

Three buyers. One product.

FOR HEADS OF AI

Ship faster, prove it.

Probari builds the assurance evidence as you build the system. Deployments stop stalling at governance review.

FOR CHIEF RISK OFFICERS

Evidence on demand.

Every use case carries a live AI Impact Assessment (AIIA), a control map and a signed sign-off chain. Auditable in minutes; defensible in front of Ofgem, the ICO or the board.

FOR DATA PROTECTION OFFICERS

DPIA and AIIA, joined up.

The intake captures lawful basis, data classes and ROPA-relevant fields once, and feeds them into both the DPIA and the AIIA. The parallel paperwork is gone.

THE PILOT

Six weeks. Two use cases. A live evidence pack.

WEEK 01

Tenant stand-up

GB Energy tenant configured for your firm. Admin, viewer and DPO accounts provisioned. The Northbrae reference deployment loaded as a working template.

WEEKS 02–04

Two live use cases

Two real AI use cases — one customer-facing, one operational — taken through intake, AIIA, control mapping and sign-off with your team.

WEEK 05

Signals feed

Operational signals wired up against the baseline. Drift triggers configured. Reassessment queue and review cadence set with the second line.

WEEK 06

Evidence pack & board read-out

Exportable evidence bundle covering both use cases across the eight-framework stack. Board-ready summary and Ofgem-ready trail walked through with the CRO.

Pilot pricing is available on application. The pilot tenant is yours to keep — there is no rebuild between pilot and production.

Book a 30-minute call.

Probari is founded and designed by Ian Curran. If you are a Head of AI, Chief Risk Officer, or DPO at a GB energy supplier, get in touch — we will book the call within five working days and walk you through the live Northbrae reference deployment.

Request a briefing

Or write directly to info@probari.uk