AI governance platform

The AI governance platform for the whole job.

Aicura is an AI governance platform for organisations in Hong Kong. It finds the AI you run, assesses every system and agent, enforces your rules where the traffic flows and watches how each one behaves, giving you the inventory, owners and controls the HKMA, SFC and PCPD expect to see.

Self-serve up to Essentials. Larger organisations, contact us.

Your AI estateGovernance moments are snapped into the Evidence Vault

Illustrative estate

Follow a scenario

AicuraMonitoring and automatic responsesYour networkOutside

Running nowMicrosoft 365 CopilotCollections assistantGitHub CopilotPeople and agents making calls

Model providersMicrosoft, OpenAI, Anthropic, AWS Bedrock, Google

Traffic, all the timeRules, when you publishData, back from your networkResponses reach back

AI register

Found by discovery or asked for by staff

Found by discoveryCMDB, GitHub, Fleet, cloud

Asked for by staffAI use requests

AI register60 systems and agents

Assessments

Risk and impact

Collections assistantHigh · 13 items

Microsoft 365 CopilotHigh · 14 items

Guardrails and gateway

Rules down to your network

Guardrails and budgetsRules v10 deployed

People and agentsMaking calls

Gateway in your networkAllows, redacts, blocks

Model providersAllowed calls only

Monitoring and automatic responses

What your network reports back

MonitoringSignals

AlertsWait on a person

Automatic responsesAt the level you choose

Oversight and trust

What you hand over

Board reportQ3

Trust CentreYour register today

ReportsPoint-in-time documents

Evidence packSealed for sharing

The platform

How Aicura works

Aicura finds the AI you run, assesses it, enforces your rules on live traffic and reports on all of it, with a sealed history of the work underneath.

AI register

Every AI system and agent you run, whether discovery found it or someone on staff asked for it. Each has a named owner. Switch the model underneath it and Aicura keeps the old version and starts a new one.

Assessments

Know what could go wrong before it does. A risk assessment rates each system, lists its risks and suggests a control for each one. An impact assessment sets out who is affected and how. Both guide your judgement.

Guardrails and gateway

Someone pastes a customer's email address into a prompt, and it is masked before it leaves your network. Your rules run on a gateway you host, which allows, redacts or blocks each request and holds each system to its budget.

Monitoring and automatic responses

An agent starts calling a tool nobody approved, overnight. An alert opens, and Aicura responds only as far as you allow, from telling the owner to acting straight away. A response can re-run the risk assessment, so production feeds back into the rating.

Oversight and trust

A committee decides on a new use of customer data, and the decision stays tied to the charter it was made under. Aicura drafts the quarterly board report from the record as it stands today, a Trust Centre page shows customers the AI you use, and the Evidence Vault holds a dated snapshot of each step.

The history. Governance decisions and enforcement actions are snapped into a sealed, timestamped history as the work happens. Nobody files anything by hand.

See the whole platform

Hong Kong

What applies to organisations in Hong Kong

Hong Kong has no AI-specific law. These are the regulator guidance and frameworks that shape how organisations there govern AI, with what each asks of you.

15 October 2026HKMA ART guidance for banksNine principles on accountability, resilience and trust, replacing the HKMA's 2019 AI principles for all authorized institutions.In forceSFC generative AI circularFour core principles for licensed corporations using AI language models, with extra safeguards and SFC notification for high-risk uses.In forcePCPD AI Model FrameworkThe Privacy Commissioner's guidance on buying and using AI under the Personal Data (Privacy) Ordinance, with risk-based human oversight.CurrentGenerative AI Guideline and Ethical AI FrameworkThe Digital Policy Office's non-binding guidance, with twelve Ethical AI Principles and an AI impact assessment template.

Where your data is hosted

Aicura is hosted on AWS in Sydney, Australia, for every customer worldwide, and the AI it uses runs on Amazon Bedrock in Australia. Your governance data is not used to train models. Transfers of your data from Hong Kong to Australia are covered by our data processing agreement.

How we handle data

AI governance for Hong Kong banks and licensed corporations

What the HKMA's 2026 ART guidance, the SFC's generative AI circular and the PCPD expect of Hong Kong banks and licensed corporations using AI.

Read the guide

Ways to start

Start on the free plan and add more as you grow

Free

Register

Find the AI already in use and give every system an owner. Pick common tools from the library and their details fill in for you. Audit log, and system and agent reports included.

Self-serve

Privacy

Review your privacy policy against the automated decision-making frameworks for your country, with suggested wording for any AI system it does not yet disclose.

Self-serve

Essentials

Know the risks of each system and what to do about them. Risk assessments with suggested controls, policies and templates, model cards that describe each system, disclosures of the decisions it makes, and guardrail rules you write.

For larger organisations

Pro and Enterprise

Put your rules on live traffic. The gateway in your network, monitoring and automatic responses, limits on what agents can do while they run, incidents, board reports and the Trust Centre.

Compare plans and pricing

Learn

Understand AI governance before you buy anything

AI governance in practiceWhat a month of AI governance actually looks like, day by day.Questions to ask about your AIWhat a board, regulator or auditor will ask, and what a good answer draws on.Get started with the free AI registerSign up, connect discovery and build your register, one plain step at a time.FrameworksThe AI laws, standards and frameworks that apply in Hong Kong, with what each requires.

Browse the guides

Questions about Aicura in Hong Kong

Does Hong Kong have an AI law?

No. Hong Kong has no AI-specific law. Organisations govern AI under existing law, most directly the Personal Data (Privacy) Ordinance, and under guidance from the PCPD, the Digital Policy Office and financial regulators such as the HKMA and SFC.

Is Aicura hosted in Hong Kong?

No. Aicura is hosted on AWS in Sydney, Australia, for every customer worldwide, and the AI it uses runs on Amazon Bedrock in Australia. Your governance data is not used to train models, and transfers from Hong Kong are covered by our data processing agreement.

Does Aicura help banks with the HKMA's AI guidance?

It supports the work the guidance describes. The AI register holds each use case with its owner, provider, model, risk rating and status, discovery surfaces AI not yet on it, and assessments, vendor records, agent controls and monitoring sit on the same records. Aicura does not assess against HKMA requirements, and the judgement stays with you.

What does Aicura cost in Hong Kong?

Prices are in Hong Kong dollars and exclude applicable taxes. The AI register is free at its smallest size, and the pricing page shows every plan and size.

Start with what you're running

Connect your sources, see what Aicura finds, and take each system from the register to enforcement at your own pace.