Crawls the web, in France

Code penal, art. 323-1 a 323-8 (Chapitre III, Des atteintes aux systemes de traitement automatise de donnees) · Read the law's page

• Do not defeat an access control, password, or other security measure to reach a French computer system; reading a public, unauthenticated page without doing so has not itself been held to violate Article 323-1.
• Do not impair, falsify, corrupt, or delete data on a French computer system in the course of an automated crawl, which Articles 323-2 and 323-3 punish separately from unauthorized access itself, with a higher penalty where the system is a State personal-data system.

Code de la propriete intellectuelle, art. L341-1 a L343-7 (Titre IV, Droits des producteurs de bases de donnees), created by loi n. 98-536 du 1er juillet 1998, transposing Directive 96/9/CE · Read the law's page

• Do not extract or reutilize a qualitatively or quantitatively substantial part of a French database's content without the producer's authorization, where the producer shows a substantial investment in constituting, verifying, or presenting that content.
• Repeated or systematic extraction or reutilization of insubstantial parts can still infringe where it amounts, in its cumulative effect, to extracting or reutilizing a substantial part.

Code de la propriété intellectuelle, arts. L331-5 to L331-10 and L335-3-1 · Read the law's page

• Do not knowingly interfere, for purposes other than research, with an effective technical measure as defined in Article L. 331-5 in order to alter the protection of a work by decoding, decrypting or any other personal intervention meant to circumvent, neutralize or remove a protection or control mechanism (Article L. 335-3-1, I).
• Do not knowingly procure or offer to others, directly or indirectly, means designed or specially adapted to defeat an effective technical measure, whether by manufacturing or importing an application, device or component other than for research, holding one for sale, loan or hire or making it available to the public, providing a service for that purpose, or inciting its use or advertising any of those processes (Article L. 335-3-1, II).
• If you are a rightholder that uses technical protection measures, take the useful steps so that they do not deprive the beneficiaries of the exceptions in point 2° of Article L. 331-28 of their effective exercise (Article L. 331-7).
• If you are a television publisher or distributor, do not use technical measures that would deprive the public of the private-copy exception (Article L. 331-9).
• Make the conditions for access to reading a work and the limits that a technical protection measure places on the private-copy exception known to the user (Article L. 331-10).

Code civil, arts. 1103, 1119 and 1171 · Read the law's page

• If you invoke general conditions against another party, bring them to that party's knowledge and obtain its acceptance, because they have effect against the other party only on those conditions (Article 1119, paragraph 1).
• Where your general conditions disagree with the other party's, treat the incompatible clauses as without effect, and let particular conditions prevail over general conditions (Article 1119, paragraphs 2 and 3).
• Perform a lawfully formed contract as you would a law, because it takes the place of law for those who made it (Article 1103).
• Do not rely on a non-negotiable clause that one party determined in advance in an adhesion contract and that creates a significant imbalance between the rights and obligations of the parties, because it is deemed unwritten (Article 1171, paragraph 1).

Code de commerce, arts. L151-1, L151-4, L151-5, L152-1 to L152-3 and L152-6 · Read the law's page

• Do not obtain a trade secret without the consent of its lawful holder through unauthorized access to, or unauthorized appropriation or copying of, any document, object, material, substance or digital file that contains the secret or from which it can be deduced (Article L151-4, point 1).
• Do not obtain a trade secret without its lawful holder's consent through any other conduct that, in the circumstances, is unfair and contrary to commercial practice (Article L151-4, point 2).
• Do not use or disclose a trade secret without its lawful holder's consent where you obtained it in the conditions of Article L151-4 or are bound by an obligation not to disclose it or to limit its use (Article L151-5, paragraph 1).

Code de la propriete intellectuelle, art. L122-5-3, created by ordonnance n. 2021-1518 du 24 novembre 2021, art. 1 · Read the law's page

• Text and data mining means applying an automated analysis technique to digital text and data to extract information such as patterns, trends and correlations.
• You may reproduce a lawfully accessed work for text and data mining, whatever its purpose, without the author's authorization, unless the author has appropriately objected, including by machine-readable means for content made available online (para. III).
• Before mining, check for that objection; an indexing or training use that ignores a machine-readable opt-out falls outside the exception and requires authorization.
• If you are a research organization, public-access library, museum, archive or film, audiovisual or sound heritage institution, or act on their behalf and at their request, you may reproduce a lawfully accessed work for text and data mining carried out solely for scientific research, without authorization (para. II).

Code de la propriété intellectuelle, art. L122-5-3 (text and data mining exceptions) · Read the law's page

• If you mine texts and data under paragraph III, make the copies or reproductions only from works you accessed lawfully (Article L122-5-3, III).
• Do not rely on the paragraph III exception for content whose author has objected appropriately, notably by machine-readable means for content made available to the public online (Article L122-5-3, III).
• Store the copies and reproductions made for mining under paragraph III with an appropriate level of security, and destroy them at the end of the mining (Article L122-5-3, III).
• If you mine under the research paragraph, do so solely for scientific research, and not where an enterprise that is a shareholder or associate of the institution carrying out the mining has privileged access to the results (Article L122-5-3, II).

Commission Nationale de l'Informatique et des Libertés (CNIL), 'AI system development: CNIL's recommendations to comply with the GDPR' and 'The legal basis of legitimate interest: focus sheet on the measures to implement in the case of data collection by web scraping' (19 June 2025), issued under Loi n° 78-17 du 6 janvier 1978 and GDPR

• Define precise data collection criteria before scraping web content to build AI training datasets.
• Exclude sensitive categories of data from web-scraped AI training datasets.
• Promptly delete irrelevant data collected through web scraping.
• Maintain an exclusion list of sources or data to omit from scraping.
• Honor site-level objections to being scraped.

Regulation (EU) 2024/1689 — directly applicable in France without national transposition; calendar amended by the Digital Omnibus, Regulation (EU) 2026/1744 of 8 July 2026 (OJ 24 July 2026, in force 27 July 2026)

• Disclose to users when they are interacting with an AI chatbot.
• Mark generative AI system outputs in machine-readable form.

Proposition de loi n° 220 (2025-2026) créant une présomption d'exploitation du contenu culturel par les fournisseurs d'IA, déposée par Mme Laure Darcos, adoptée par le Sénat le 8 avril 2026, transmise à l'Assemblée nationale le 9 avril 2026