Asimov AI
Asimov AI is an artificial intelligence platform that analyses legislative acts and amendment files and monitors parliamentary activity. Built in Rome for public institutions, companies and trade associations, it is reachable only through a credential request.
What is Asimov AI?
Asimov AI is an artificial intelligence platform specialised in the analysis of legislative documents: complex legal acts, and the amendment files that accompany them through Parliament. It is published by ASIMOV AI S.R.L., an Italian innovative startup founded in Rome in March 2024 by Ennio Filicicchia, a computer engineer trained at ETH Zurich, and Federico Bonomi, a legislative professional with experience in the Senate and at Oxford.
The vendor builds its pitch on four claims. The platform is presented as reliable, combining automated and manual checks; in-depth, because it autonomously reconstructs the meaning of provisions in light of the cited legal framework, working from official databases; fast, handling hundreds of provisions in minutes and saving more than 90% of the time a manual analysis would take; and analytics-oriented, with advanced analysis tooling. The English pages claim accuracy of over 98% and describe the coverage as the European legislative domain, where the Italian pages state both differently.
Behind the interface sit modules for acts, amendments, provisions, parliamentary iters, digests, keywords, recurring questions, quotations and monitoring, the last including a calendar, external legal sources and regional sources. A dedicated panel attaches Corte Costituzionale decisions to a given amendment, a dossier brief groups the material of a project, and amendment files can be imported straight from a Senato URL. Eight notification streams, from analysis completed to iter status and Gazzetta keyword matches, are delivered by immediate email, daily digest or in-app message. Workspaces let several members share one monitoring profile. The underlying data comes from Normattiva, the Camera dei Deputati, the Senato della Repubblica, the Corte Costituzionale and the Gazzetta Ufficiale.
The institutional anchoring is unusual for so young a company. In July 2024 Asimov AI won the Camera dei Deputati call for artificial intelligence applied to parliamentary work, inside the GENAI4LEX consortium led by the University of Bologna, and it lists a scientific committee of six academics from Bar-Ilan, Sapienza, LUISS, Bologna and IUSS Pavia. Poste Italiane, La Repubblica and SOGEI are displayed as clients. The vendor itself warns that its AI processing may produce errors, and that the Normattiva data it uses is not the official version of the texts.
What it does
- Analyse complex legislative acts and amendment files, in any format
- Reconstruct the meaning of a provision in light of the legal framework it cites, on official databases
- Produce comparative and consolidated versions of the texts a provision modifies
- Query a finished analysis through a chatbot and advanced filters
- Monitor parliamentary iters, the legislative calendar and regional sources
- Deliver notifications and scheduled digests on the activity being followed
- Group custom keywords to filter documents and surface the ones that matter
When to use Asimov AI / When not to
A quick filter to help you decide if Asimov AI is the right fit.
When to use Asimov AI
- Parliamentary and legislative officers who have to work through amendment files containing hundreds of provisions
- Public affairs and government relations teams tracking parliamentary iters, decree-laws and regional legislative sources
- Regulatory and compliance analysts who need the consolidated text and the legal framework a provision actually cites
- Public bodies and administrations already working on official sources such as Normattiva, the Camera dei Deputati and the Senato
- Large organisations, SMEs and trade associations whose business depends on following Italian regulatory production
When not to use Asimov AI
- Anyone working outside the Italian legislative perimeter: the corpus is Normattiva, the Camera, the Senato, the Corte Costituzionale and the Gazzetta Ufficiale
- Lawyers looking for a general-purpose legal assistant for contracts, case law research or client advice
- Buyers who need to compare prices before talking to a salesperson: there is no published price, no free trial and no self-service sign-up
- Developers who want to embed the analysis in their own systems: no public API documentation exists on the site
- Users who need a mobile application, or an interface in a language other than Italian or English
How to use Asimov AI
A typical end-to-end flow, from setup to results.
- Fill in the credential request form; the Asimov team gets back to you to activate the access, as there is no self-service sign-up
- Log in to the web platform, with nothing to install and no client software to deploy
- Describe the monitoring profile of your organisation, its activities and the regulatory areas to follow, shared with the members of your workspace
- Upload legislative acts and amendment files, in any format, or import an amendment file directly from a Senato URL
- Start the automatic analysis
- Review the results: provisions, normative references, comparative texts and consolidated texts
- Query the analysis through the chatbot and the advanced filters
- Set recurring questions so that a downloadable report is generated at the end of every analysis
- Register recurring amendments so they are compared against each newly uploaded file
- Organise keywords into groups, tune the notification streams and channels, schedule digests, and set the interface and notification language to Italian or English
Pros & Cons
Pros
- Verifiable institutional anchoring: winner of the July 2024 Camera dei Deputati call for AI applied to parliamentary work, inside the GENAI4LEX consortium led by the University of Bologna
- A scientific committee of six named academics from Bar-Ilan, Sapienza, LUISS, Bologna and IUSS Pavia, and six research publications listed by the vendor
- Named official data sources rather than an unspecified corpus: Normattiva, Camera dei Deputati, Senato della Repubblica and Corte Costituzionale
- Claimed time saving of more than 90% against manual analysis, with accuracy stated above 98% on the English pages
- Signatory of the European Commission's AI Pact and listed in the EU Transparency Register
- Explicit commitment that uploaded documents, including publicly available ones, are never shared with other users
- An AI disclaimer shown inside the product itself, not only buried in a legal page
Cons
- No public pricing whatsoever: no pricing page, no range, not one named plan
- No terms of use are published; the only legal documents are a privacy policy and a cookie policy, and the privacy policy covers the showcase website rather than the platform
- No free trial and no self-service path: access is granted only after a credential request processed by the team
- The Italian and English versions of the site contradict each other on accuracy, above 95% against over 98%, and on scope, Italian against European legislative domain
- No public API documentation, no mobile application and no browser extension
- Training of models on customer data is never addressed anywhere, and the list of processors is available only on request
- A very young and very small company: founded in 2024, EUR 10,000 of share capital, with no social account or corporate profile published
Pricing & Plans
No price is published anywhere on the website: there is no pricing page, no price range and no named plan, and neither a free plan nor a free trial is announced. The vendor states only that it offers several plans varying according to the features required, and directs prospects to its sales team, which places the tool squarely in a contact-sales model. The subscription tiers are said to be adapted from SMEs to large corporations, public bodies and trade associations, but none of them is quantified. The only monetary figures available concern the company itself, such as its EUR 10,000 share capital, and these are corporate records rather than prices. Obtaining any figure therefore requires requesting a quotation.
Data, GDPR & hosting
A consolidated view of how Asimov AI handles your data.
GDPR overview
GDPR implementation is documented and specific. The privacy notice is drafted under Articles 13 and 14 GDPR and Italian Legislative Decree 196/2003, and was last updated on 30 June 2025. It sets out the rights of access (Art. 15), rectification (Art. 16), erasure (Art. 17), restriction (Art. 18), portability (Art. 20), objection (Art. 21) and withdrawal of consent (Art. 13(2)(c)), a one-month response deadline extendable to three months in complex cases (Art. 12), and the right to complain to the Garante per la protezione dei dati personali. As a general rule no personal data is transferred outside the European Union; where a transfer occurs, Article 46 safeguards apply. A data protection officer answers at dpo@asimov.law. The company also claims full compliance with the EU AI Act, signed the European Commission's AI Pact and appears in the EU Transparency Register under number 609402093820-54. The notice covers the website only.
Who owns the data?
The only policy published governs the website www.asimov.law, not the application platform, and it was last updated on 30 June 2025. The controller is Asimov AI S.R.L., Via del Gambero 19, 00187 Rome, VAT number 17556441008, reachable through dpo@asimov.law. Data collected by the site is handled only by staff authorised under Article 29 GDPR and Article 2-quaterdecies of the Italian Privacy Code, or by processors appointed under Article 28, and the notice states that personal data is never disseminated to undefined recipients, including by simple availability or consultation. For documents uploaded into the platform, the single commitment published is a FAQ answer: they are never shared with other users.
Reuse rights
No terms of use are published, so nothing on the site states what a client may do with the analyses, consolidated texts and reports the platform produces; reuse rights are left to whatever contract is negotiated with the vendor. What the privacy policy does describe is the company's own use of website data, across six declared purposes: browsing the site, informational and commercial communications, answering enquiries, handling job applications, allowing data subjects to exercise their rights, and legal defence. The legal bases invoked are consent (Article 6(1)(a)), legitimate interest (Article 6(1)(f)), contract and pre-contractual steps (Article 6(1)(b)) and legal obligation (Article 6(1)(c)). The categories collected are first name, surname, email address, telephone number, career and education history and skills, together with browsing data such as pages visited, volumes, session identifiers, IP address, URLs and cookies. The notice adds that if you are only browsing the website, this is the only processing performed on your personal data. No page collected mentions any use of customer data to train models.
Data retention & training
Hosting summary
The vendor declares no hosting country and no hosting region. The only geographic statement in the privacy policy is a transfer clause: as a general rule the controller does not transfer personal data outside the European Union or to international organisations, and where a transfer does occur it is placed under the safeguards of Article 46 GDPR. That is a commitment about transfers, not a declaration that the platform is hosted in the European Union, and the two should not be confused. AWS appears on the about page as a partner under its startup programmes, which says nothing about where the application runs. The cookie policy names Cloudflare, jsDelivr, unpkg and Webflow as third-party services of the website. One network observation, which is not a vendor declaration: the showcase site resolves to 34.154.186.117, an address operated by Google LLC in Milan, Italy, and that describes the marketing site rather than the platform where client documents are processed. The FAQ states that uploaded documents are stored securely and encrypted, without naming a location or a provider. Any organisation with a data residency requirement should obtain the answer in writing before contracting.
Things to keep in mind
Risks and trade-offs to weigh before adopting Asimov AI.
- The site contradicts itself from one language to the other: the English pages claim accuracy above 98% and a European legislative domain, while the Italian pages of the same FAQ claim above 95% and an Italian legislative domain. The fourth product block changes too, Analytics in English against Funzionale in Italian, and the mention of the Italian public procurement marketplace exists only in Italian. Two versions of the same claims, and no published methodology behind either figure.
- The vendor's own disclaimer states that amendment processing and rule summaries are generated by AI systems that may produce errors or inaccuracies, and that the Normattiva data used is not the official version of the texts, the only official text being the printed Gazzetta Ufficiale. An output read too quickly can be mistaken for a legal certainty it never claimed to be.
- Deskilling risk: a tool that rebuilds the sense of a provision in minutes makes it tempting to stop reading the underlying norm. On amendments, the cost of a missed cross-reference is paid downstream by the institution, not by the software.
- No terms of use are published, and the privacy policy covers the showcase website rather than the application. Nothing contractual on the site governs the documents clients upload; the only commitment is a FAQ answer, which is a weak instrument if a dispute ever arises.
- Model training on customer data is never addressed anywhere on the site, and the list of processors is only available on request. Confidential parliamentary or corporate material deserves a written answer on both points before the first upload.
- AWS, Innova S.R.L., LUISS and EUSAIR are displayed as partners, not as certifiers or integrations, and no security certification such as ISO 27001 or SOC 2 is claimed anywhere. Partner logos should not be read as an audit.
- Corporate information around the tool is inconsistent: the incorporation date is not established, with two aggregators publishing conflicting dates under contradictory labels, one of them still showing a former address, and the site's own llms.txt is out of date against the site it describes.
Setup & Integrations
Technical difficulty
Low on the technical side, moderate on the organisational side. Nothing is installed: the platform runs in a browser, with no API key to place and no integration to build. The real gate is access itself, granted only after a credential request handled by the vendor's team. Once the account is open, the value depends on configuration work that belongs to the users rather than to IT: describing the organisation's monitoring profile, building keyword groups, setting recurring questions and recurring amendments, choosing among the notification streams and scheduling digests.
Deployment
Integrations
Supported languages
Behind Asimov AI
Resources
All the official URLs gathered for verification and reference.
Frequently asked questions
What does Asimov AI actually do?
How reliable are the analyses?
Are the documents I upload shared with anyone else?
How much does it cost?
How do I get access?
Which languages does the platform support?
Is there a mobile application?
Which data sources does it work on?
Can the AI get things wrong?
Should you pick Asimov AI?
Asimov AI is a vertical tool and makes no attempt to be anything else. It is built for people who follow legislative production professionally: parliamentary staff, public affairs teams, regulatory analysts, legal officers in public bodies and large organisations. On that ground it is convincing. The corpus is named and official, the analysis goes beyond summarising and tries to rebuild the meaning of a provision against the rules it cites, and the monitoring layer of iters, calendars, regional sources, notification streams and digests matches the way that work is actually done.
Its credibility is also unusual for a company founded in 2024: a Camera dei Deputati call won in July of that same year, a consortium led by the University of Bologna, a named scientific committee, a signature on the European Commission's AI Pact, and three sizeable clients on display.
The reservations are commercial and contractual rather than technical. Nothing about the price is public, no terms of use exist, and the only legal document published covers the showcase website rather than the platform where documents are actually uploaded, so the promise that files are never shared between clients rests on a FAQ answer. The site also contradicts itself between its Italian and English versions, both on accuracy and on whether the scope is Italian or European. Access requires a credential request, with no trial to test the claims beforehand.
Worth the conversation if legislative monitoring is part of your job and you are equipped to negotiate the contractual terms the website does not publish. Anyone looking for a general-purpose legal assistant should look elsewhere.
- Choosing a selection results in a full page refresh.
- Opens in a new window.