Universal Registration Document 2023

4. Corporate Social Responsibility Non-Financial Performance - AFR

4.3.8 Responsible marketing: unique approach

4.3.8.1 Ethical principles applied to all business lines
1) Confidentiality

Respect for the confidentiality of client data and projects is a fundamental value. It is required from 100% of employees, in addition to the obligations undertaken by employees in their employment contract with the Groupe. Teams may have access to sensitive information; they are now systematically asked to sign specific confidentiality agreements (NDAs – non‑disclosure agreements). Intellectual property, whatever the type of creation or output, is also protected. Experts in trademark law or copyright or database law (data base), within the legal teams, must be consulted well upstream of projects. Data protection and security specialists must also be involved in all projects to ensure that these issues are addressed strictly.

2) Intellectual property

As a creative company, Publicis Groupe has always been committed to respecting and protecting intellectual property, an increasingly complex topic to manage in a digital and ultra‑connected world and with images created through artificial intelligence. It is in this spirit that the PMX Digital team has set up an exclusive contract with WIPO (World Intellectual Property Organization) to identify and exclude sites that violate intellectual property.

3) Artificial intelligence

The use of artificial intelligence, or algorithms based on deep learning (also called machine learning), is already integrated into the Groupe’s business lines, with interesting and successful advances in Media activities, in data with Epsilon and in digital transformation with Publicis Sapient. This makes it possible to meet very large‑scale customization needs. For several years, the Groupe’s Responsible Marketing policy has defined five key principles to be respected:

  1. equity and inclusion: by applying the “Inclusion by design” principle to combat unconscious bias and ensure a diversity of viewpoints in the team so as not to offend anyone;
  2. reliability and security: critical review is part of the process of designing an IT program to ensure clarity and completeness for each user;
  3. privacy protection and data security: the “Privacy by design” principle is applied to these projects, as described in the Publicis Groupe data protection policy;
  4. transparency and accountability: it is imperative to monitor performance to identify irregularities and continue to learn from all experiences;
  5. “tests and trials”: tests are the crucial steps to ensure that the project complies with regulations and industry best practices, and to validate the effectiveness of the proposed campaign or technological solution.

In February 2024, the Chair of the Management Board, accompanied by the Directoire+, presented to all employees the developments related to the arrival of generative artificial intelligence (AI) in the Groupe’s business lines. The main lines of the corporate deployment project around CoreAI and the evolution towards an "Intelligent System" (.is) were laid out. The use of machine learning (automated deep learning) has already been integrated into many activities for several years, in creative but also at Epsilon and Publicis Sapient. The transformation brought about by generative AI must be based on clear ethical rules, particularly in terms of transparency, loyalty, responsibility, confidentiality and security.

The Groupe has set up an operational governance for artificial intelligence (AI), in view of its strategic needs in its various business lines, and the interest expressed by clients. In the short term, these various tools, particularly those authorized by the Groupe or expressly requested by the client, need to be appropriated in order to understand their use in a holistic way, integrating transparency on the sources of learning and on the data collection mechanisms. Internal uses of what AI has generated, whether text, images, sound or video, must comply with these principles. Employee received the first key policies related to AI, including Generative AI Responsible Use Guidelines and Generative AI Legal Guidelines. They have access to several tools on Marcel to test them in an internal and protected environment, to learn how to use them effectively and to collaborate on projects. All corporate functions legal, HR, IT, CSR, etc. are committed to contributing to this transformation.

4) Lobbying practices

Some assignments may involve lobbying and strategies to influence decision‑makers. The lobbying teams operate in compliance with the laws and the Groupe’s rules, particularly concerning combating conflicts of interest and anti‑corruption. In accordance with legal obligations and best practices, the teams concerned are identified and, where required, their activities as representatives of interests are declared, both in terms of the Transparency Register of the European Parliament and the European Commission, or on a country‑wide basis, listed in the digital repertoire of representatives of interests managed by the High Authority for the Transparency of Public Life in France (HATVP), and in the United States, where the rules of the Lobbying Disclosure Act apply, or where this relates to the FARA (Foreign Agent Registration Act), with registration in compliance with the subjects and organizations concerned.