evidoria

← Back to browse

Good practice Imported

GENAI4LEX-B — AI-assisted legislative drafting and compliance-checking for Italy's Chamber of Deputies

Italy · Rome · See the Italy profile · See the Rome profile

Evidence: Descriptive / self-reported Top 83% 40/100 · Ask Evidence Copilot about this practice

Italy's Chamber of Deputies selected GENAI4LEX-B, a University of Bologna-led consortium's hybrid AI tool, from 28 competing proposals to help draft legislation and check bills against existing law — still a pilot with no published outcome data.

Details

Maturity
Pilot
Promoter
Camera dei Deputati, with a University of Bologna (ALMA-AI)-led consortium
Period
{'end': '', 'start': 'July 2024'}
Keywords
generative AI, legal informatics, legislative drafting, parliamentary support

Context

In 2024, Italy's Chamber of Deputies (Camera dei Deputati) ran a Call for Interest for proposals to use generative AI in parliamentary work, receiving 28 proposals from 15 institutions; on 25 July 2024 it selected three winners, with GENAI4LEX-B winning the category for support in drafting parliamentary initiative acts. The consortium is led by Prof. Monica Palmirani at the University of Bologna's ALMA-AI centre, with LUISS Guido Carli, the Universities of Turin and Verona, CNR's Institute of Legal Informatics and Judicial Systems (Cnr-Igsg), and spin-offs Aptus.AI and Asimov AI. It combines EuroVoc legal ontologies and the Akoma Ntoso/LegalDocML legislative-document standard with a deductive-logic engine, aiming to help MPs and staff locate regulatory sources, summarise committee amendments, and check draft-bill consistency with existing law and constitutional-court rulings.

Results

As of the most recent reporting available, GENAI4LEX-B remains a pilot: no time-savings, accuracy or usage figures have been published. Chamber officials have stated that 'the human component must be preserved in all operational and decision-making processes.' A peer-reviewed paper in Parliamentary Affairs has separately raised concerns about the neutrality, transparency and legitimacy of introducing AI into legislative procedure, and about its effect on the informal, uncodified dimensions of parliamentary work.

Implementation
Indicative cost
Medium (€50k–€500k) — Multi-institution university/CNR consortium plus two commercial spin-offs; no published budget figures.
Time to results
Medium (1–3 years) — Selected 25 July 2024; still described as a pilot in the latest available reporting, with no rollout date announced.
Staffing & skills
Led by Prof. Monica Palmirani at the University of Bologna's ALMA-AI centre, with LUISS Guido Carli, the Universities of Turin and Verona, CNR's Institute of Legal Informatics and Judicial Systems (Cnr-Igsg), and spin-offs Aptus.AI and Asimov AI

Conditions for success

  • Selected competitively from 28 proposals across 15 institutions via a formal Chamber of Deputies call
  • Built on open legal-technology standards (Akoma Ntoso/LegalDocML, EuroVoc) intended for reuse across legislative bodies
  • Chamber officials have publicly committed to preserving human decision-making in the process

Common failure modes

  • No published time-savings, accuracy or usage data over a year after selection
  • A peer-reviewed Parliamentary Affairs paper raises unresolved concerns about the neutrality, transparency and legitimacy of introducing AI into legislative procedure

Where it fits

Governance type
national parliament
Scale
pilot (single legislative chamber)
Income level
high-income

Commonly funded by

National / regional programmes

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Similar practices you may find useful