Residents’ data in consumer tools
Pasting an objection or a care notification into a free chatbot takes seconds. Afterwards, nobody can explain where that data went.
Local government
AI for your organisation. With residents’ data under your own control. And accountability that holds.
Public bodies may use generative AI, but under stricter conditions than companies: a risk analysis in advance, reliable models and, under the AI Act, a fundamental rights impact assessment and registration for high-risk use. werqly runs that AI on a node that belongs to your organisation alone, with roles per department and an audit log that feeds your accountability.
- Residents’ data stays in your own infrastructure
- Open-weight models, including European ones
- Audit log per user and per source
Running on your own node
Role: Permits
- Mistral Large
- Llama 70B
- Qwen 32B
Safeguards active
Insight: anonymous (default)
- National ID detection
- Case numbers blocked
werqly Node — your own AI node, in your own server room
The risk
Where it goes wrong in the public sector.
No overview of what is running
AI features arrive inside software you already bought. Without a register you do not know which applications exist, or which will count as high-risk.
Accountability you cannot reconstruct
For a council question, a freedom of information request or an audit, you have to show what happened. Without logging, that is detective work.
Benefits
What it brings your organisation.
More room for the work that serves residents, without their data ending up at an external AI service.
Faster through applications and mail
Applications, objections and incoming mail are classified and summarised, so case handlers reach the point sooner.
Policy and regulation findable
By-laws, policy documents and earlier decisions searchable with source references, instead of scattered across folders.
Accountability from the system
Who used which model with which source is recorded. That saves work on risk analyses, registers and council questions.
No dependence on one vendor
Open-weight models on your own node: switching models is not a migration project.
Use cases
What public bodies use it for.
Automations are built per organisation, usually following the TIMA Scan. The decision always stays with a civil servant.
Preparing applications and objections
Summarising documents and checking them against policy, with source references, for the case handler to assess.
Questions from residents
Classifying incoming questions and preparing a draft reply from approved information.
Council and executive papers
First drafts of memos, answers to council questions and summaries of lengthy annexes.
Searching policy
By-laws, policy rules and earlier decisions searchable for those authorised to see them.
Preparing information requests
Finding and ordering documents and flagging passages that need assessment, before a lawyer looks at them.
Minutes of meetings
Decisions and action points from recordings, processed on your own node.
AI security and governance control
AI security and governance, set up for a public body.
werqly AI Control gives the CISO, the DPO and information management one place to limit, log and account for AI use.
Roles per department
Permits, social services, taxes and communications each get their own sources, models and automations.
Safeguards on ID and case numbers
National ID numbers, case numbers and other patterns are blocked per role before they reach a model.
Audit log to your SIEM
Every prompt, every model and every automation run, with a timestamp, exportable to your own monitoring.
Insight without looking over shoulders
Anonymous by default: trends per department, never traceable to an employee. Per user only with two-person approval.
Model provenance recorded
Which model version processed which document can always be traced. The basis for your AI register and the AI Act.
Automations under management
Every automation has an owner, a version and a defined set of sources. Switching one off is immediate.
Regulation
What your frameworks look for.
werqly does not take over your obligations. It does keep the processing inside your own infrastructure and let you show what happened.
| Framework | What it asks | How werqly fits |
|---|---|---|
| National policy on generative AI | Many countries set their own conditions for public bodies. The Dutch government-wide position of April 2025, for example, allows wider use after a risk analysis, with reliable models and a preference for European, open-source and in-house solutions. | A dedicated node with open-weight models fits those conditions. The risk analysis and the choice per application stay with your organisation. |
| AI Act (Regulation (EU) 2024/1689) | Public bodies deploying high-risk AI assess the impact on fundamental rights in advance and register that use in the EU database. AI literacy and transparency duties apply to all AI use. | werqly does not make an application compliant by itself. Recorded model provenance, logging and roles do support your documentation and oversight. |
| GDPR | A legal basis, purpose limitation and appropriate security for residents’ data, and a DPIA where the processing is likely to be high risk. | Processing on a node that belongs to your organisation alone, with no external AI vendor in the processing. Safeguards block ID and case numbers per role. |
| Algorithm registers | Several countries ask public bodies to publish their high-impact algorithms, such as the Dutch national Algorithm Register. | Your AI register and the audit log provide what such a publication needs: purpose, data, model and who is responsible. |
| Freedom of information, archiving and security baselines | Documents about an administrative matter may be subject to information requests, records must be kept, and information security follows the baseline for public bodies. | You set logging and retention periods yourself. The node sits in your own network, under your own segmentation and monitoring. |
This page is not legal advice. Always assess the use of AI with your own data protection officer, compliance function or professional body.
Recommended setup
Which set-up suits a public body.
For municipalities and public agencies we recommend your own server room: residents’ data is then processed entirely in-house, and the node falls under your own information security. Organisations without a suitable server room can choose werqly Cloud: a dedicated node in a locked rack in our European datacentres.
- A node in your own server room, never shared
- Roles per department, so not everyone sees every source
- With self-management, your own IT department operates the node
Calculate your private AI costs.
Enter your work email and number of employees. We work out four setups for you straight away: in your own server room or in our datacentre, self-managed or with werqly Pro management. werqly is not a cheap AI subscription: you pay for your own hardware and full control.
- 1Enter your email and number of employees
- 2Compare four proposals
- 3Request a quote straight away
Working out the proposals…
Your own server room
Your node in your own server room or rack
werqly Cloud
No server room? A locked rack in our EU datacentre
werqly Pro management
Your own server room
We operate the node and the models
— per month
—
- Data never leaves your premises
- No outside party in the processing
- Updates, monitoring and support by werqly
werqly Pro management
werqly Cloud
We operate the node and the models
— per month
—
- For organisations without a server room
- Dedicated node in a locked rack, never shared
- Updates, monitoring and support by werqly
Self-managed
Your own server room
Your IT department operates it; we supply software and updates
— per month
—
- Data never leaves your premises
- No outside party in processing or management
- Full control with your IT department
Self-managed
werqly Cloud
Your IT department operates it; we supply software and updates
— per month
—
- For organisations without a server room
- Dedicated node in a locked rack, never shared
- Your IT department runs policies and automations
Indicative amounts, excluding VAT. Custom automations are quoted separately. Only the quote is binding.
FAQ
Questions from public bodies.
May civil servants use generative AI?
That depends on national policy. The Dutch government-wide position of April 2025, for instance, gives more room than its 2023 predecessor but sets conditions: a prior risk analysis, reliable models and responsible use. You assess each application against them, which is simpler with an environment of your own.
Does werqly receive our residents’ data?
No. Prompts, documents and results are processed on your organisation’s node. The management link to AI Control carries configuration, updates and only the insights you release yourself, never the content of cases or files.
Does werqly help with registers and the AI Act?
werqly does not make an application compliant by itself, but it does supply what you need: which model, which sources, which roles and what use. That comes from AI Control and the audit log, instead of from asking around departments.
Can we release different sources per department?
Yes. In AI Control you link groups from your identity provider to models, sources and automations. Social services then see different sources than permits or communications.
Further reading in the knowledge base
Generative AI in local government: what is allowed, and how to set it up
Public bodies may use generative AI, under conditions. What the AI Act asks of them, what the Dutch 2025 policy shows, and how to set up AI in-house.
Read the article →AI in permit decisions: faster handling, without giving up the reasoning
Permit decisions run on short statutory deadlines. Where AI speeds up the work of case handlers, and where the decision has to stay with the authority.
Read the article →AI in social services: where it becomes high-risk, and what is still possible
AI that decides who is entitled to a benefit or service is high-risk AI. What that means, and which applications remain justifiable for public bodies.
Read the article →Other sectors
Law firms and notaries
It is not about AI, it is about professional secrecy. Case files and client data never leave your firm.
View the sector →Healthcare
Patient data, national hosting rules and medical confidentiality: AI that runs in your own server room.
View the sector →Financial sector
Client data, DORA and the AI Act: AI whose location you can evidence.
View the sector →