Governed AI for Regulated Industries

Built for the
stakes you
carry.

Every regulated firm faces the same underlying problem — AI adoption without governance creates liability, evidential gaps, and regulatory exposure. AiSensa is built to solve it, whatever your sector.

Azure UK South UK Data Residency Guaranteed ISO 42001 Aligned Built for Regulated Industries
AI Governance Risk — Without AiSensa

A fee earner uploads a confidential client document marked "Restricted — Not for AI Processing" to an ungoverned AI tool. The breach has already happened.

1
Data Residency Outside UK jurisdiction
2
Client Confidentiality Absolute duty breached
3
Regulatory Exposure No audit trail
4
PI Insurance Unquantifiable liability
5
Duty of Care Professional obligation unmet

Legal

Confidential client matter data is an existential governance risk.

A single breach of client matter confidentiality can result in regulatory sanction, professional liability, and reputational damage that takes years to recover from. For legal firms handling vulnerable client data — Court of Protection, mental capacity, family — the stakes are higher still.

Your fee earners are already using AI. Not because they are careless — but because AI is genuinely useful and the governed pathway either doesn't exist or creates too much friction. AiSensa solves both problems at once.

SRA Regulatory Exposure

The SRA's approach to AI is evolving rapidly. Firms without documented, evidenceable governance are exposed to regulatory scrutiny they cannot defend against. A policy document is not a defence. An audit trail is.

ICO & UK GDPR Liability

Uploading client personal data to an ungoverned AI tool — even unintentionally — is a data breach under UK GDPR. The ICO's enforcement activity around AI data handling is accelerating.

PI Insurance & Panel Retention

Professional indemnity insurers and panel procurement teams are beginning to ask for evidence of AI governance controls. Firms that cannot provide it are exposed at renewal and at tender.

Client Data Sovereignty

Clients are beginning to ask where their matter data goes when AI is used on it. A firm that can answer that question with a verified, architectural guarantee is ahead of the market.

Scenario — Court of Protection Practice

The governance didn't change
what the firm believes.
It changed what the firm can prove.

!

The Problem

"The problem isn't Sarah."

Sarah Mitchell, a trusted Court of Protection fee earner, uploads a Deputyship Order marked "Strictly Confidential — Not for AI Processing" to an ungoverned AI tool. The tool flags the sensitivity. The breach has already happened. No record. No audit trail. No defensible position. The problem isn't Sarah — it's the absence of governance at the point where AI meets the work.

01

Discover

"AiSensa already knew."

A single Graph API subscription indexes the entire M365 tenant — SharePoint, Teams, OneDrive. Sarah's COP/2024/0047 files were fingerprinted at ingest. Every document carries an invisible Custom XML fingerprint — content hash, governance record, timestamped — that survives edits, renames, and re-uploads. The compliance scan surfaces 671 documents with zero governance classification. Policy exists. The estate doesn't reflect it. Now it can.

02

Govern

"Close the gap before it becomes a crisis."

The Court of Protection — Protected Person policy is applied. AI_Classification and AI_Usage_Constraint metadata is written back to SharePoint in real time — enforced at the point where documents meet AI tools, not at the cloud boundary. An attempted upload of a welfare assessment is blocked immediately — clear message, event logged, defensible record created.

03

Enable

"Governance that works with people, not against them."

Sarah queries the Deputyship Order. The document requires supervisor approval — routed to Richard Okafor, the supervising partner named in the order itself. He reviews the full query and matter context. He approves. The AI delivers a legally grounded response — COP Rules 2017, PD 9E, MCA 2005 s.4, three relevant cases. Timestamped. UK-resident. On the record. Sarah's productivity is intact. The firm's compliance record is intact.

04

Monitor

"The operational picture, over time."

Four weeks after deployment: 73% AIG-001 compliance, up from 0%. 671 documents fingerprinted. 47 governed AI interactions. 6 supervisor approvals on the record. Sarah's full event timeline visible in a single audit trail. When an insurer, a regulator, or a client asks for evidence of AI governance controls — the answer is a timestamped, immutable record. Not a policy document.

How It Works in Practice

Three governed states.
Every AI interaction, resolved.

AiSensa doesn't block AI. It ensures every interaction with governed content follows the right pathway — automatically.

Human in the Loop

Supervisor Approval

Required before AI proceeds

High-sensitivity documents require named supervisor review before any AI interaction is permitted. The query, the matter context, and the approval decision are all logged.

Sarah queries a Deputyship Order. Routed to Richard Okafor — supervising partner. He approves. The AI responds. Everything on the record.
Permitted

Approved Pathway

Automatic — no friction

Documents classified as approved pathway proceed automatically. No approval required, no delay. The interaction is logged. Sarah's productivity is intact.

Sarah selects a CoP precedent classified as Approved Pathway. AI proceeds automatically. Fast. Logged. Fully compliant.
Prohibited

Hard Block

AI access prevented

Welfare assessments, vulnerable adult records, and documents carrying a hard prohibition are blocked at the point of AI interaction. Clear message. Directed to supervising partner. Logged — not to penalise, but to protect.

Sarah selects a welfare assessment. Hard stop. Clear message. Attempt logged. Client data never reaches an AI system.
MCA

The Mental Capacity Act is a legislative expression of human-in-the-loop thinking. AiSensa is its technological counterpart. When a Court of Protection document is flagged before AI interaction, we are not imposing a governance constraint. We are expressing the firm's own professional values — in code, in workflow, in the moment of decision.

What You Get — Legal

Everything a legal firm needs to govern AI responsibly.

AiSensa is shaped around the specific regulatory and workflow context of UK legal practice. Not a generic AI platform with a legal use case bolted on.

Matter-level AI controls

Governance applied at the matter and document level — not just the user level. The right controls follow the right documents, wherever they go.

Policy connected to SRA & ICO obligations

AI-assisted policy creation built from your actual regulatory obligations. Tested for enforceability. Updated in minutes when an incident occurs or obligations change.

Compliance evidence for PI insurance

A live, improving governance record that answers due diligence questions from PI insurers at renewal and panel procurement teams at tender. Compounds over time as a competitive asset.

UK data sovereignty — architectural guarantee

Azure UK South and UK West only. No data leaves the UK. Not a configuration option — a hard architectural fact you can communicate directly to clients and regulators.

Fee earner productivity — research & drafting

Governed AI agents for legal research, document drafting, and precedent work. The same AI capability your fee earners want — with the governance the firm requires.

Seneca Legal — 4 Weeks Post-Deployment

671
Documents fingerprinted
73%
AIG-001 compliant
(up from 0%)
47
Governed AI interactions
6
Supervisor approvals on record

From 0% to 73% compliance in four weeks — using existing SharePoint infrastructure, no new Microsoft licences, and a read-only discovery phase that required no changes to the live estate.

The Competitive Landscape

AiSensa wins by doing what others cannot do at all.

Harvey.ai is the reference point. £3,000 per seat per year. No UK data residency guarantee. No governance infrastructure. AiSensa is the answer for every firm that needs governed AI but cannot justify — or access — Harvey's price point.

AiSensa
Harvey.ai
UK data residency — architectural guarantee
✓ Azure UK South / UK West
✗ Not guaranteed
Governance infrastructure (policy, controls, monitoring)
✓ Full AI Control Lifecycle
✗ Not provided
Document fingerprinting & endpoint governance
✓ Fingerprint + Desktop + Browser Agents
✗ No endpoint presence
Works on existing Microsoft infrastructure
✓ Microsoft native via Graph API
Partial — separate upload required
Accessible to specialist & mid-market firms
✓ Impact-first, no upfront platform commitment
✗ £3,000/seat, £30k+ minimum
Human-in-the-loop governance by design
✓ Architectural — every interaction
✗ Automation-first model
Evidenceable compliance for regulators & insurers
✓ Audit report, management report, dashboard, alerts
✗ Not the product's purpose

Next Step — Legal

See AiSensa against
your live estate.

A scoped proof of concept — read-only, no changes to your infrastructure, no data leaving your environment. We map your estate, surface your exposure, and show you exactly what governed AI looks like in your firm.

Read-only discovery phase — no changes to your live estate, no data leaving your infrastructure.

Accountancy

AI adoption without governance is a compliance liability and a missed opportunity at the same time.

Accountancy firms handle highly sensitive client financial data at volume. Fee earner time directly drives revenue — and the pressure to adopt AI to accelerate that work is real. But AI tools used outside a governance framework create both regulatory exposure and reputational risk that firms are only beginning to understand.

The good news: the same governance framework that manages the risk also unlocks the productivity. AiSensa doesn't make you choose between compliance and capability.

FRC & ICAEW Obligations

The FRC's evolving guidance on AI in audit and the ICAEW's AI framework both require firms to demonstrate how AI is being used, governed, and overseen in client work. Documentation is not enough. Evidence is.

ICO & UK GDPR Liability

Client financial data is highly sensitive personal data under UK GDPR. Any AI processing — even by a fee earner using a consumer tool — must be lawful, documented, and auditable. The ICO is watching.

Client Confidentiality

Accountancy firms carry the same fundamental duty of client confidentiality as legal firms. A fee earner uploading a client's management accounts to an ungoverned AI tool is a breach — regardless of intent.

Audit Quality & Independence

AI-assisted audit work must be documentable and defensible. If the AI interaction isn't logged, it didn't happen — or worse, it happened in a way you cannot account for.

What You Get — Accountancy

Governed AI that earns its keep.

AiSensa enables accountancy firms to adopt AI productively and safely — governed from day one, evidenceable to regulators and clients, and built around the workflows fee earners actually use.

Client data governance at document level

AI controls applied at the client file and engagement level. The right governance follows the right documents — wherever fee earners take them.

Policy connected to FRC & ICAEW obligations

AI governance policy built from your actual regulatory obligations — FRC, ICAEW, ACCA, ICO. Tested for enforceability. Updated when obligations change.

Credit control & debtor prioritisation

Governed AI agents that accelerate high-value back-office work — identifying priority debtors, drafting client communications, and surfacing actionable insight from aged debt data.

Fee earner productivity — within governed boundaries

The same AI capability your fee earners want for research, drafting, and client work — with the governance controls the firm requires. No trade-off between speed and compliance.

Audit-ready compliance evidence from day one

A complete, independent audit trail for every AI interaction. Exportable for FRC inspection, ICAEW review, or client due diligence — from the moment you go live.

Next Step — Accountancy

Start with a real problem.
See governed AI deliver.

We identify the right use case for your firm — a governance gap, a productivity bottleneck, a compliance obligation — and demonstrate measurable value before you commit to a platform.

Fixed-price discovery sprint. Problem-first. Value before platform commitment.

Financial Services

Coming soon.

AiSensa's governance architecture is built for FCA-regulated environments — customer data protection, audit trails for AI-assisted decisions, and UK data residency as a client-facing guarantee. We're working with our first financial services clients now.

FCA Regulatory Context Customer Data Protection UK Data Residency Guarantee EU AI Act Aligned