Artificial Intelligence

The development and deployment of artificial intelligence systems raise issues of regulation, intellectual property, data protection and liability. The phased application of the EU Artificial Intelligence Act (AI Act) and the rise of generative AI require companies to secure their technologies, contracts and practices.

The Artificial Intelligence department of d&a partners advises AI model providers, software publishers, startups, integrators, investors and players in the creative industries on the development, commercialization and use of AI systems.

The firm acts upstream of projects, in audits, in contract negotiations and in litigation. Its approach combines legal expertise, an understanding of technology and knowledge of the standardization work applicable to AI.

AI Act: compliance with the EU Artificial Intelligence Act

Regulation (EU) 2024/1689, known as the AI Act, applies to providers, deployers, importers and distributors of AI systems, as well as to providers of general-purpose AI models (GPAI). Each company must identify its role in the value chain, qualify its systems and determine the obligations that actually apply to it.

The firm advises in particular on:

• the legal qualification of AI systems according to their risk level, their purpose and the role of each player;
• compliance: transparency, documentation, human oversight, high-risk AI systems;
• the obligations of GPAI model providers, including for models presenting a systemic risk;
• internal policies, AI use charters and governance procedures;
• the interplay with the GDPR, sector-specific regulations and European and international technical standards.

Compliance frameworks are designed in proportion to each company’s activities, risks and operational constraints.

Legal audit of AI models and training data

The provenance of training data, licenses, third-party rights and the terms of use of pre-existing models may jeopardize the development or exploitation of an AI system.

The firm’s legal audits cover in particular:

• data sources and collection, training and fine-tuning operations;
• intellectual property rights and rights reservations under the text and data mining exceptions;
• the legal bases for processing personal data;
• licenses and restrictions on use, including for third-party models, open source components and models accessed via API;
• documentation and transparency obligations.

The objective: identify the risks and define the legal, contractual or organizational measures needed to control them.

Generative AI, copyright and intellectual property

The automated generation of text, images, videos, voices, software or avatars raises questions of rights ownership, originality and the use of pre-existing works, in particular in the audiovisual, music, publishing, video game and advertising industries.

The firm advises developers of generative AI solutions, rights holders, producers, creators and user companies on:

• the protection of AI-assisted creations and rights in generated content;
• the terms of use of AI tools, exploitation licenses and infringement risks;
• digital avatars, virtual influencers, AI-generated characters and synthetic voices, at the crossroads of copyright, neighboring rights, image rights and personal data;
• code produced with AI assistance (AI-assisted coding, vibe coding): ownership of rights, open source licenses and securing software development.

AI contracts, technology licenses and SaaS solutions

AI projects rely on contractual chains between model providers, software publishers, integrators, service providers and user companies. These contracts must organize rights in technologies, data and generated content, and allocate liability.

The firm drafts and negotiates in particular:

• development and integration agreements;
• model licenses and SaaS agreements;
• terms of use and terms of sale;
• API access agreements and technology partnerships;
• R&D agreements, consortia and transfers or acquisitions of AI technologies.

Particular attention is paid to ownership of outputs, rights to input and output data, intellectual property warranties, confidentiality and information security.

Personal data protection and artificial intelligence

The GDPR applies at each stage of the life cycle of an AI system, from the building of training datasets to the use of its outputs. The firm advises on:

• legal bases and the allocation of responsibilities between parties;
• data protection impact assessments (DPIA) and international transfers;
• the exercise of data subjects’ rights;
• the risks specific to generative AI: memorization or regurgitation of data by models, biometric recognition, synthetic content reproducing a person’s image or voice.

This support combines GDPR requirements, the recommendations of the CNIL (the French data protection authority) and the obligations of the AI Act.

Governance, ethics and assessment of AI systems

Beyond compliance, AI governance requires control, documentation and oversight mechanisms adapted to the risks. The firm assists companies and organizations (international organizations, ministries, public services) with:

• defining AI governance policies and drafting ethics charters;
• framing professional uses and organizing internal responsibilities;
• the legal assessment of algorithmic risks: transparency, bias, discrimination, human oversight, traceability, explainability.

This practice draws on technical standardization work (ISO/IEC standards, European work of CEN-CENELEC) to align legal obligations with operational risk management methods.

AI-related litigation and liability

The firm assists companies, developers, professional users and rights holders in disputes relating to:

• the unauthorized use of works or data to train models;
• the reproduction of protected content;
• providers’ contractual warranties and difficulties in operating AI solutions;
• damage caused by automated systems.

It works in coordination with its intellectual property, contracts, regulatory and commercial litigation teams.

Legal, academic and institutional expertise

The practice is led in particular by Matthieu Quiniou, partner admitted to the Paris Bar, Doctor of Law and Associate Professor (maître de conférences) at Université Paris 8. He is in particular:

• a member of the UNESCO expert network on artificial intelligence and the rule of law;
• an expert for the European Commission for the Efficiency of Justice (CEPEJ) of the Council of Europe, for which he coordinates the Artificial Intelligence Advisory Board;
• a member of the AI standardization committee at AFNOR;
• involved in research, training and publication on the legal and societal implications of digital technologies.

This combination of legal practice, research, European regulation and standardization enables the firm to support projects involving complex legal and technological issues.

Frequently asked questions on artificial intelligence law

Which companies must comply with the AI Act?

Companies that develop, place on the market, import, distribute or deploy AI systems in the European Union, under the conditions set out in the regulation. Obligations vary according to the company’s role, the type of system, its uses and its risk level. Providers of general-purpose AI models are subject to specific obligations.

How is a legal audit of an AI model carried out?

It reviews the provenance of training data, intellectual property rights, licenses, personal data processing, the terms of use of third-party models and regulatory obligations. Its scope depends on the technical architecture of the system, its uses and the way it is commercialized.

Is AI-generated content protected by copyright?

Under French law, protection requires an original creation attributable to a natural person. Content generated entirely by AI cannot in principle be protected. A creation made with the assistance of an AI may be protected if the human contributions are sufficiently original, which requires keeping and organizing evidence of the creative process.

Which lawyer should you consult to develop or commercialize an AI system?

A lawyer practicing artificial intelligence law advises providers and publishers of AI systems on regulatory compliance, intellectual property, training data and commercialization agreements. At d&a partners, Matthieu Quiniou, partner, handles these matters together with the firm’s other teams.

Topics: Artificial intelligence · AI law · AI Act · EU Artificial Intelligence Act · AI compliance · AI legal audit · AI models · Generative AI · Intellectual property · Copyright · Training data · GPAI · Fine-tuning · LLM · GDPR · AI contracts · SaaS · Model licenses · API · Algorithmic governance · Digital avatars · Deepfakes · Vibe coding · AI litigation · AI standardization

Are you developing, commercializing or using artificial intelligence systems?

The Artificial Intelligence department of d&a partners can help you structure your project legally, identify your obligations and secure your technologies.

Contact our team for an initial discussion.