March 29. 2024. 7:08

The Daily

Read the World Today

French data protection authority lays out action plan on AI, ChatGPT


The French data protection watchdog, the National Commission on Informatics and Liberty (CNIL), published an action plan on Tuesday (16 May) addressing privacy concerns related to Artificial Intelligence, particularly generative applications like ChatGPT.

ChatGPT, the world’s most famous chatbot, expanded its user base to 100 million users in the first two months after its release. While its popularity grows, so do concerns over how it gathers and processes personal data.

“In the face of recent news on artificial intelligence, and in particular so-called generative AIs such as ChatGPT, the CNIL publishes an action plan for the deployment of AI systems that respect the privacy of individuals.” reads the announcement.

However, the decision of the Garante opened the door to a fragmented approach across the EU. The European Data Protection Board, which gathers all EU data regulators, established a task force to ensure consistent enforcement.

CNIL’s role

At the European level, the French authority is already considered one of the most influential in the EU, meaning the plan could shape how Europe’s regulators approach ChatGPT and similar technologies.

Domestically, a source informed on the matter told EURACTIV under conditional anonymity that the French data protection authority is positioning itself to lead the national enforcement of the AI Act, a landmark EU legislation to regulate AI based on its capacity to cause harm.

This ambition is enshrined in the action plan that states that “this work will also make it possible to prepare for the entry into application of the draft European AI Regulation.”

Italian data protection authority bans ChatGPT citing privacy violations

The Italian privacy watchdog mandated a ban on the popular chatbot ChatGPT and launched an investigation on its provider OpenAI for suspected breaches of EU data protection rules.

Four-step approach

The four steps of the action plan consist in understanding the technology, guiding its development, creating an AI ecosystem and controlling AI systems.

The first step will be dedicated to answering questions related to data protection, such as transparency on the training datasets, protecting publicly available data from scraping, safeguards from biases, and the inputs provided by the users.

As these aspects are prioritised in the EU and France, the CNIL dedicated internal resources to answer these questions and has already published a dossier sharing its vision on answering data protection issues.

The “guiding AI development” stream wishes to guide generative AI companies towards technical developments respecting personal data by publishing guideline documents and sharing best practices and rules.

The third “ecosystem” stream has three layers. It aims to expand the regulator’s existing regulatory sandbox to innovative AI-based projects.

CNIL also launched a competition to assist companies in their compliance with European data protection rules. In addition, the French authority launched a project for providers of “enhanced video surveillance” in the context of the French government’s experimentation of these tools provided in the law on the Paris Olympic and Paralympic Games in 2024.

The last stream of CNIL’s action plan touches upon its core competencies: audit and control of digital systems. The privacy watchdog will focus on compliance with the use of enhanced video surveillance, the use of AI in the fight against fraud and investigate complaints lodged on generative AI.

On this last point, the French regulator notes that it has already received several complaints against OpenAI, on which it will coordinate with the European task force.

European data protection authorities launch task force on ChatGPT

The European privacy regulators decided on Thursday (13 April) to launch a dedicated task force to address the privacy concerns related to the world’s most famous chatbot.

Read more with EURACTIV

Majority of EU countries support scanning of audio communications

Majority of EU countries support scanning of audio communications