Legal Foundation
Art. 13 of Law September 23, 2025, n. 132 (Official Gazette n. 223 of 25/09/2025)
Paragraph 1: "The use of artificial intelligence systems in intellectual professions is aimed solely at the exercise of instrumental activities and support to professional activity and with prevalence of intellectual work object of the work performance."
Paragraph 2: "To ensure the fiduciary relationship between professional and client, information relating to artificial intelligence systems used by the professional is communicated to the recipient of the intellectual service with clear, simple and exhaustive language."
Effective date: October 10, 2025 - The disclosure is mandatory for all intellectual professionals who use AI in the exercise of their activity.
The tool has been designed to ensure compliance with paragraph 2, requiring all information necessary to communicate clearly, simply and exhaustively the use of AI to the client.
Why This Data Is Required
1. Professional/Firm Data (Section 1)
Legal rationale: The disclosure must allow the client to clearly identify the AI processing controller and whom to contact to exercise their rights. Complete identification data (name, address, contacts, professional register) guarantee traceability and professional responsibility.
Mandatory fields:
- Name/denomination - identifies the controller
- Professional type - demonstrates membership in intellectual professions (recipients of L. 132/2025)
- Complete address - necessary for physical identification of the office
- Email - contact channel for exercising client rights
2. AI Usage Level (Section 2)
Legal rationale: Paragraph 1 of Art. 13 requires that AI be used only for "instrumental and support activities" with "prevalence of intellectual work". Declaring the usage level allows the client to understand the intensity of AI support and verify that human work always remains prevalent.
- Basic: demonstrates minimal use (organization only)
- Intermediate: limited operational support
- Advanced: substantial support but always with prevalent human supervision
3. AI Systems Used (Section 3)
Legal rationale: Compliance with the AI Act (EU Reg. 2024/1689) requires identification of AI systems, providers and purposes of use. This allows the client to:
- Know which tools are used on their data/documents
- Evaluate the reliability of providers
- Verify consistency between declared purposes and actual practices
Why supported activities are mandatory: Specifying which activities (research, drafting, translation, synthesis) are supported by AI demonstrates that the professional maintains control over critical activities reserved for human judgment.
4. Supervision and Quality Control (Section 4)
Legal rationale: The requirement of "prevalence of intellectual work" (Art. 13 paragraph 1) is implemented through:
- Supervision level: demonstrates that every AI output is verified by a human being
- Control procedures: describes how quality is guaranteed (source comparison, double checks, regulatory verifications)
- Excluded activities: lists what AI can never do (e.g. strategic decisions, ethical evaluations, official certifications) - this protects the core of the intellectual profession
Without these elements, the client could not verify that human intellectual work remains prevalent.
5. Client Rights (Section 5)
Legal rationale: The "fiduciary relationship" referred to in paragraph 2 implies that the client can:
- Refuse the use of AI for their practice
- Request complete human control over any output
- Know when a document was generated/assisted by AI
- Exercise opt-out from AI processing of their data
The ability to quickly select all rights facilitates the adoption of a maximalist approach to client protection.
6. Professional Responsibility (Section 6)
Legal rationale: Art. 13 does not attenuate professional responsibility. The explicit declaration that "responsibility remains entirely with the professional, regardless of AI use" is fundamental to:
- Avoid disputes about dilution of responsibility
- Protect the client in case of AI errors
- Maintain deontological standards of the profession
How to Use the Tool
Initial Wizard
When opening the page, the wizard asks you to choose the AI usage level. This choice automatically pre-fills some fields (audit frequency, supervision level) according to best practices for that level. You can always modify these values.
Quick-Add Buttons
For the most common AI systems (ChatGPT, Claude, Gemini, Copilot) the quick-add buttons automatically insert name, provider, version and typical purposes. You must still select the specific supported activities for your use.
Email Auto-copy
The professional email entered in Section 1 is automatically copied to the "Contact person for AI information" field (Section 8), avoiding duplication.
Quick Rights Selection
The "Select all rights" button in Section 5 allows you to quickly select/deselect all client rights. It is recommended to keep them all for maximum protection.
Generation and Export
The tool generates the document in real time. You can:
- Preview: view the complete document before downloading
- Direct export: download in TXT, PDF, DOCX, HTML formats even without opening the preview
Available formats:
- TXT: simple text, compatible with any editor
- PDF: ready for printing and delivery to the client
- DOCX: editable Word format for customizations
- HTML: for publication on the firm's website
Privacy and Security
Client-side architecture: The tool operates exclusively in the user's browser. No data is:
- Sent to external servers
- Saved in databases
- Transmitted to third parties
- Stored after closing the page
This design choice guarantees maximum confidentiality of professional and client data entered in the disclosure.
Validation and Compliance
The tool has been designed to ensure regulatory compliance:
- Art. 13 L. 132/2025: All information required by paragraph 2 is present
- AI Act (EU Reg. 2024/1689): Identification of systems, providers, purposes, human supervision
- Clear and simple language: The template uses terminology understandable even to non-experts
Important Disclaimer: This tool generates a draft compliant with regulatory requirements. The professional remains responsible for verifying, customizing and updating the document based on their specific professional situation. Legal review is recommended before use.
Frequently Asked Questions
Is the disclosure mandatory even if I use AI only for internal activities?
Yes. Art. 13 paragraph 2 does not distinguish between internal and external use. If AI supports the professional service in any way, the client must be informed.
What happens if I don't provide the disclosure?
Failure to provide disclosure may constitute a violation of Art. 13, with possible deontological and professional liability consequences. Furthermore, it compromises the fiduciary relationship required by the regulation.
Can I use AI for strategic decisions if I review them?
Paragraph 1 requires "prevalence of intellectual work". Strategic decisions must remain predominantly human. AI can provide analytical support, but the final decision must be the result of intellectual elaboration by the professional.
Do I need to update the disclosure every time I change AI system?
Yes. Any substantial modification (new system, new purposes, change in supervision level) requires updating and re-communication to the client.
Can the client refuse the use of AI?
Yes. The right of refusal is implicit in the fiduciary relationship. If the client refuses, the professional must evaluate whether they can perform the service without AI or whether they must decline the assignment.