Skip to content

AI governance your board can defend, line by line.

Aegis interviews every department, scores your AI governance against Ghanaian law and international standards, and produces signed policy documents. Every number traces to an answer, a document and a clause of law.

Now enrolling pilot organisations for 2027

Why is data governance at level 2?Illustrative example
The rules engine applied the cap. The AI wrote none of these numbers.

Pilot programmeEnrolling a small group of Ghanaian organisations for 2027.

Led by RAMSYSLegal and governance practitioners review every engagement.

Built in GhanaGhanaian law first, with a sovereign Ghana-hosted option.

What you get

A score the law is allowed to overrule.

Most governance tools average their way to a number, so a serious legal breach disappears into a good overall result. Aegis scores each dimension from the answers, and then applies the obligations that bind this organisation. Where one is unmet, it sets a ceiling the rest of the score cannot climb past.

A board sees the ceiling the law imposed and the clause that imposed it — not just a low number somebody will argue about.

How the scoring works

2.0of 5, organisation-wide maturity
Maturity by dimension, illustrative
DimensionMaturity barLevel
Strategy and leadership3.5
Data governance and privacy2.0 (capped by a rule)
Security2.0 (capped by a rule)
Risk and compliance1.0
People and AI literacy2.5
Technology and tooling3.0
Use-case management2.0

Capped at 2.0Not registered as a data controller with the Data Protection Commission. The rules engine applied the ceiling; nothing else this organisation does can lift it until the registration exists.

Illustrative data. Hatching marks the maturity a binding legal gap removed. No model set any of these numbers.
April 2026

Ghana launched its National AI Strategy. Boards are now being asked how their organisation governs the AI it already uses.

The problem

Most AI governance reviews end in a report whose conclusions nobody can re-derive six months later.

Scores that depend on who asked

Two reviewers interviewing the same department reach different conclusions. A board cannot defend a number it cannot reproduce.

Findings with no trail

A finding in a PDF does not point to the answer or document behind it, so an auditor has to take it on trust.

Law mapped by hand, every time

The mapping from Ghanaian law to international standards is rebuilt for each engagement instead of maintained once.

How it works

From interview to a document anyone can verify.

The AI does the interviewing and the drafting. Rules set the scores. People review and sign. Each step leaves a record the next one depends on.

  1. Interview

    An AI interviewer talks to each department about the AI it actually uses, including chat assistants and tools adopted without approval.

  2. Evidence

    Departments upload the policies, registers and contracts that back their answers. Each document is attached to the answer it supports.

  3. Score

    A deterministic rules engine scores seven dimensions of maturity and records a finding against every control. The AI does not set any score.

  4. Review and sign

    RAMSYS consultants review the findings and the drafted documents, then your signatories approve and sign them.

  5. Verify

    Every issued document carries a QR code. Anyone holding a copy can confirm it is genuine and still current.

Explore the platform

What you receive

Three signed documents, not a slide deck.

Each one is assembled from a clause library vetted by lawyers, filled in from your findings, approved by RAMSYS, signed by your executives and issued with a QR code that confirms it is genuine.

For the board and executive committee

Executive AI Readiness Scorecard

One maturity level for the organisation, seven dimensions beneath it, and the binding obligations still unmet.

Signed, versioned and verifiable

One for each department interviewed

Departmental AI SOP Playbooks

Standard operating procedures written for how that department actually uses AI, with named owners.

Signed, versioned and verifiable

For signature by your accountable executives

Master Corporate AI Policy

An organisation-wide AI policy assembled from a lawyer-vetted clause library, not written freshly by a model.

Signed, versioned and verifiable
Who it is for

One assessment, read four ways.

The same evidence serves the board, the compliance team, IT and the public interest. Each sees the part they are accountable for.

Boards and executives

A maturity level you can defend, with the binding obligations listed separately from good practice.

Boards and executives: How it helps

Compliance and legal

Findings mapped to Act 843 and Act 1038, with the answer and document behind each one.

Compliance and legal: How it helps

IT and data

A register of the AI tools in use across the organisation, including the ones nobody approved.

IT and data: How it helps

State-owned enterprises

Sovereign hosting in Ghana and documents that stand up to public scrutiny.

State-owned enterprises: How it helps

Frameworks

Ghanaian law first, international standards alongside.

Findings say whether a control is a legal obligation or good practice. A gap against Act 843 is not presented with the same weight as a gap against a voluntary framework.

FrameworkHow Aegis treats itWhat it covers in an assessment
Data Protection Act, 2012 (Act 843)GhanaBinding in Ghana
  • Registration as a data controller
  • Lawful basis and purpose limitation for AI processing
  • Data subject rights, including decisions made about people by automated means
Cybersecurity Act, 2020 (Act 1038)GhanaBinding in Ghana
  • Incident reporting procedures
  • Access control over AI tools and the data they reach
  • Security of systems that host or call AI services
ISO/IEC 42001 — AI management systemsInternational standardBenchmark
  • AI policy and assigned roles
  • An inventory of AI systems in use
  • Risk and impact assessment
NIST AI Risk Management FrameworkUnited States, voluntaryBenchmark
  • Governance and accountability structures
  • Mapping AI use and its context
  • Measuring and managing identified risks
EU Artificial Intelligence ActEuropean UnionBinding only where output reaches the EU
  • Human oversight of high-risk systems
  • Transparency to people affected by AI
  • Risk classification of AI use cases

Read how each framework is applied

Software and people

Software that scales, with experts who stand behind it.

Aegis is operated by RAMSYS, a Ghanaian legal and consulting firm. The platform does the repeatable work so practitioners can spend their time on judgement.

About RAMSYS advisory

The platform

  • Interviews every department
  • Collects and links evidence
  • Scores against the control library
  • Drafts from vetted clauses
  • Tracks adoption and actions

RAMSYS practitioners

  • Scope the engagement
  • Review every binding finding
  • Edit and approve the documents
  • Brief the board
  • Advise on remediation
Trust and security

Your governance data is evidence. We treat it that way.

The controls below are part of how the platform is built, not settings you have to remember to turn on.

Isolated by the database

Row-level security scopes every query to one organisation, even if application code is wrong.

Hosted in Ghana

The sovereign tier keeps your data in Ghana, for organisations that need it to stay there.

An audit trail you cannot edit

Significant actions are written to an append-only log that records who did what, and when.

Verifiable documents

A public check confirms a document was issued and is current, without revealing its contents.

AccraBanjulLagosNairobiCairoJohannesburgAddis Ababa

Ghana first. The Gambia next. The continent after that.

Every assessment is scored against the law that actually binds the organisation taking it, which means a country is not a translation — it is a library of instruments, obligations and crosswalks that RAMSYS’s lawyers verify before a single client sees it.

Where the data may not leave the country, the same platform runs in country, on your own infrastructure.

Read about security and data residency

Pilot programme 2027

Help shape AI governance for Ghanaian organisations.

We are enrolling a small number of organisations for 2027. Pilot organisations receive a full assessment and work directly with the team building the platform.

What pilot organisations receive

  • A full assessment of the departments you choose
  • The scorecard, department playbooks and corporate AI policy, signed and issued
  • A named RAMSYS consultant throughout
  • A board briefing on the findings
  • Direct input into what the platform does next
Rollout

Signed documents in about 90 days.

A typical engagement for a handful of departments. Larger organisations run in waves on the same plan.

  1. Day 30Assess

    Scope agreed and departments named. Interviews run with each department, evidence is collected, and the AI tools in use are registered.

  2. Day 60Score and draft

    The rules engine scores all seven dimensions. RAMSYS reviews every binding finding, and the scorecard, playbooks and policy are drafted.

  3. Day 90Sign and adopt

    Documents are approved, signed and issued with QR verification. Adoption tracking starts and remediation actions have owners and dates.

See the engagement plan

Questions

What people ask first.

Does the AI decide our score?

No. The AI conducts the interviews and drafts prose. Every score and every finding comes from a deterministic rules engine applied to your answers and evidence, so the same inputs always give the same result.

Each number links back to the answers, documents and controls that produced it, and a RAMSYS consultant reviews the findings before anything is issued.

We do not build AI models. Is this for us?

Yes. Most organisations use AI through chat assistants, office software and vendor tools, often without a formal decision to adopt them. Aegis finds that use through interviews, so it does not need access to a model pipeline or a data science team.

Which laws and frameworks are covered?

Ghana’s Data Protection Act, 2012 (Act 843) and Cybersecurity Act, 2020 (Act 1038) are binding. ISO/IEC 42001 and the NIST AI RMF are used as benchmarks. The EU AI Act is treated as binding only for AI systems whose output reaches the EU; elsewhere it is a benchmark.

Where is our data stored, and who can see it?

Each organisation’s data is isolated by the database itself using row-level security, so a query cannot return another organisation’s records. A sovereign deployment keeps your data hosted in Ghana. See Trust for the detail.

Does this replace our lawyers?

No. Aegis supports legal advice and does not replace it. Documents are assembled from a clause library reviewed by lawyers, and RAMSYS practitioners review each engagement, but your organisation should still take advice on its own circumstances.

What does a pilot involve?

A full assessment of an agreed set of departments over roughly 90 days, run with a RAMSYS consultant, ending in signed documents. Pilot organisations also shape the product through regular feedback sessions. Ask about the pilot programme.

See your organisation’s readiness the way your board will.

A 45-minute walkthrough with a RAMSYS consultant: the interview, a scored assessment, and the documents it produces.