2026-10-29 - AI and the professions in Legislative Decree 179/2026: mandatory training for professional orders and fair compensation adjusted to risk
This contribution continues the series block devoted to
Legislative Decree no. 179 of 7 October 2026
, following the analysis of the
. For the professions, the decree does not merely promote initiatives: it imposes obligations on professional orders, sets deadlines for adjusting their rules, and introduces a new rule on fees. These provisions have the most direct impact on practitioners, including the legal profession.
Mandatory training within professional orders
Initial vocational training and continuing training courses run by professional orders “include” literacy and training courses on the use of artificial intelligence systems (Article 47(1)). The wording is mandatory, and the minimum content is set by law (Article 47(2)):
- technical aspects: functioning, querying and instruction techniques, potential and limits of the systems used in the relevant professional field;
- legal aspects: the rules governing AI systems, with particular regard to the AI Act and the national implementing provisions;
- ethical and professional conduct aspects: the professional’s responsibilities when using AI systems, information obligations towards the client and respect for the human-centric principle in Article 1(1) of
.
The courses provide for a minimum number of hours for each continuing training assessment period under the rules of each category; content, delivery methods and frequency are determined by the national councils of the orders and colleges, or by the bodies that regulate training for each category, through the adjustment of the rules provided for in paragraph 5, in compliance with these minimum requirements (Article 47(3)). The decree does not set the number of hours: it leaves that to each category.
Who must adjust and by when
Within six months of entry into force, by 23 April 2027, the national councils of the orders and colleges, or the bodies regulating professional training for the respective categories, must adjust their rules or organizational acts on initial and continuing training to include AI courses (Article 47(5)). The adjustment follows each category’s procedures, including, where provided, the supervising authority’s opinion or approval.
The same rules apply to courses promoted by the umbrella associations of non-regulated professions under Article 3 of Law 4/2013, where the use of AI is relevant to the field of activity (Article 47(4)). Under Article 5 of the same law, these associations make the courses and arrangements for verifying actual compliance with the obligation publicly known (Article 47(5)).
Professional orders or trade associations may carry out training activities aimed at professionals, including under agreements with universities, public research bodies, and AFAM institutions (Article 43(2)).
The health professions
For the health professions, the State-Regions Conference establishes the inclusion of courses in continuing medical education programs by agreement, to be concluded within six months in compliance with the minimum content and hours (Article 47(6)). Within the same time limit, the National Commission for Continuing Education sets the percentage of the three-year training obligation to be devoted to AI, and Agenas identifies the programs to be included in management training courses (Article 51(3) and (4)).
In my view, the two provisions will need careful coordination: the agreement under Article 47 and the percentage under Article 51 operate on the same continuing medical education obligation, with the same deadline and different actors.
Fair compensation adjusted to risk
Article 49 introduces a new rule. A professional’s use of AI systems may trigger an adjustment to fair compensation under Law 49/2023, based on parameters that take into account the risk classification of the system used under the AI Act (Article 49(1)).
Within twelve months, that is, by 23 October 2027, three groups of decrees must be supplemented with the parameters for the adjustment (Article 49(2)):
- the decrees adopted under Article 9 of Decree-Law 1/2012, on the parameters for regulated professions;
- the decree provided for in Article 13(6) of Law 247/2012, on the parameters for lawyers;
- the decrees adopted under Article 1(1)(c) of Law 49/2023.
The provision implements the delegation criterion in Article 24(2)(f) of Law 132/2025, which provides for fair compensation adjustable based on the responsibilities and risks connected with the use of AI systems. The decree does not indicate the direction of the adjustment: it does not say whether the use of a system should increase or reduce the fee, only that the parameter must take the risk class into account.
Concluding remarks
For professional orders, the 23 April 2027 deadline is close, and in many categories the adjustment requires a step before the supervising authority. For practitioners, AI training becomes part of the ordinary training obligation, with ethical content that directly concerns information obligations towards clients.
On fair compensation, I think the reading most consistent with the delegation links the adjustment to the responsibilities assumed: using a system classified as high-risk entails greater obligations and responsibilities, and the fee should reflect that. However, AI systems often reduce the time needed to perform a service, and the decrees on parameters will have to decide whether and how to account for this as well. Until they are supplemented, the rule has no applicable parameters.
For those wishing to navigate the regulatory framework referred to in this contribution, the
of this site are available: the
, the
and the
.
The reference sources are available on the institutional websites:
Legislative Decree no. 179 of 7 October 2026
,
and
.
New to Gemini? This capsule is best read with a Gemini client, such as Lagrange.