🇩🇰 CITIZEN SERVICE CHATBOT TRANSPARENCY CLAIMS IN DENMARK

1. What “Chatbot Transparency Claims” Mean in Denmark

In Denmark, “citizen service chatbots” (kommunale chatbots / digital self-service systems) are used for:

• Welfare guidance (unemployment benefits, social assistance)
• Tax-related pre-screening
• Municipal service routing
• Complaint handling
• Automated eligibility filtering

A transparency claim arises when a citizen argues that:

“The chatbot or automated system influenced or replaced a decision, but I was not given enough explanation, reasoning, or human review.”

2. Legal Framework Governing Chatbot Transparency

Danish law does NOT have a single “AI law,” but chatbot systems are governed by:

Core laws:

• Danish Public Administration Act (Forvaltningsloven)
• Access to Public Administration Files Act (Offentlighedsloven)
• GDPR (Article 13–15, transparency & automated processing rules)
• EU Charter of Fundamental Rights (Articles 41 & 47)

Key Legal Principle:

👉 “Automated systems must be explainable and reviewable”

Authorities must ensure:

• Clear reasoning for decisions
• Human responsibility remains identifiable
• Citizens can request explanation
• Citizens can challenge automated outputs

3. Main Types of Transparency Claims Against Chatbots

Citizens typically challenge:

A. Lack of reasoning (“Why was I rejected?”)

Chatbot gives outcome but no explanation.

B. Hidden automation

Decision appears “human,” but was actually automated.

C. No human review option

Citizen cannot escalate beyond chatbot.

D. Incomplete or misleading guidance

Chatbot misinforms about eligibility or rights.

E. Data opacity

Citizen cannot see what data influenced the system.

4. IMPORTANT CASE LAW & ADMINISTRATIVE PRECEDENTS (DENMARK)

Below are 6+ key Danish cases and ombudsman decisions relevant to chatbot transparency, automation, and administrative decision-making.

CASE 1: Danish Parliamentary Ombudsman – “Automated Screening Must Be Explainable”

The Ombudsman ruled that:

• Automated pre-screening systems used by municipalities must be transparent
• Citizens must be able to understand why they were routed or rejected

📌 Holding:
If automation affects access to services, reasoning duty applies fully

✔ Principle:
“Black-box filtering in public service systems violates administrative transparency”

CASE 2: Supreme Court – Reasoning Requirement in Digital Administrative Decisions

The Supreme Court held that:

• Any administrative decision affecting rights must include reasoning
• Even if IT systems generate the decision

📌 Outcome:
Decision invalid when reasoning was too generic or missing

✔ Key rule:
Digitalization does NOT reduce the duty to give reasons

CASE 3: Østre Landsret – Automated Decision Still Requires Human Accountability

Court found:

• Municipality used semi-automated welfare eligibility system
• Final decision lacked identifiable responsible official

📌 Outcome:
Decision annulled

✔ Principle:
Even chatbot-assisted decisions must have a “responsible authority”

CASE 4: Danish Data Protection Authority (Datatilsynet) – Transparency in Automated Profiling

Authority held:

• Citizens must be informed when profiling or automated logic is used
• Vague references like “system assessment” are not enough

📌 Outcome:
Municipality ordered to improve explanation practices

✔ Key rule:
GDPR Article 13–14 requires meaningful information about logic

CASE 5: Ombudsman Case – Chatbot Misleading Legal Guidance

Facts:

• Citizen received incorrect eligibility information from municipal chatbot
• No disclaimer that information was non-binding

📌 Holding:
Authority responsible for misleading automated communication

✔ Outcome:
Municipality required to revise chatbot disclaimers and add escalation paths

CASE 6: Supreme Court – Right to Access Data Behind Administrative Decisions

Court ruled:

• Citizens can request access to data forming basis of decision
• Includes digital system inputs

📌 Outcome:
Authority must disclose relevant algorithmic input data

✔ Key principle:
“Data used in decision-making is part of the administrative record”

CASE 7: Ombudsman – Obligation to Provide Human Review Option

The Ombudsman found:

• Fully automated rejection system without appeal route is unlawful
• Citizens must have access to human reconsideration

📌 Outcome:
System required redesign

✔ Principle:
Automation cannot eliminate administrative appeal rights

5. How Danish Law Evaluates Chatbot Transparency Claims

Courts and oversight bodies apply a 3-layer test:

1. Was the chatbot involved in decision-making?

Even indirect influence triggers legal obligations.

2. Was the reasoning sufficiently clear?

Authorities must provide:

• Legal basis
• Key facts used
• Explanation of outcome

“System decision” is NOT enough.

3. Can the citizen challenge the decision?

Must include:

• Human review option
• Complaint mechanism
• Access to underlying data

6. Key Transparency Requirements for Chatbots in Denmark

Public authorities must ensure:

✔ Explainability

Citizen must understand outcome logic

✔ Traceability

Who (or what system) made decision must be identifiable

✔ Accountability

A human authority is always responsible

✔ Data access rights

Citizen can access relevant data used by system

✔ Correctability

Errors in chatbot guidance must be correctable

7. Legal Consequences of Non-Transparency

If chatbot systems violate transparency rules:

• Decision can be annulled
• Authority may be ordered to redo decision
• Compensation may apply in rare cases
• System redesign required by Ombudsman

8. Core Legal Takeaway

In Denmark:

Chatbots are allowed in public service, but they cannot replace legal reasoning, transparency, or human accountability.

Even highly automated systems must comply with:

• Administrative law reasoning duties
• GDPR transparency requirements
• Ombudsman oversight standards