European UnionAI Act, Article 59 (further processing of personal data for developing certain AI systems in the public interest in an AI regulatory sandbox)
Regulation (EU) 2024/1689, Article 59
In force since .
An AI governance rule binding public and private bodies.
- Private right of action
- No
- How the hook was established
- express
- What makes it apply
- Market targeting, Operator establishment, Place of effect
- Obligation class
- Governance, Security
- Audit expectation
- none
As of .
What it requires
- This permission applies from , the Regulation's general application date: Article 113, as amended by Regulation (EU) 2026/1744, gives Chapter VI, which holds Article 59, no date of its own. For a high-risk AI system related to a product covered by the Union harmonisation legislation listed in Section B of Annex I, Articles 57, 58 and 59 apply only in so far as the requirements for high-risk AI systems under this Regulation have been integrated in that Union harmonisation legislation (Article 2(2), as replaced by Regulation (EU) 2026/1744).
- If you take part in an AI regulatory sandbox, you may process personal data lawfully collected for other purposes solely for the purpose of developing, training and testing certain AI systems in the sandbox, when all of the following conditions are met (Article 59(1)).
- The AI systems are developed for safeguarding substantial public interest, by a public authority or another natural or legal person, in one or more of these areas: public safety and public health, a high level of protection and improvement of the quality of the environment, energy sustainability, safety and resilience of transport systems and mobility, critical infrastructure and networks, and efficiency and quality of public administration and public services (Article 59(1), point (a)).
- The data processed are necessary for complying with one or more of the requirements referred to in Chapter III, Section 2, where those requirements cannot effectively be fulfilled by processing anonymised, synthetic or other non-personal data (Article 59(1), point (b)).
- There are effective monitoring mechanisms to identify whether any high risks to the rights and freedoms of the data subjects, as referred to in Article 35 of Regulation (EU) 2016/679 and Article 39 of Regulation (EU) 2018/1725, may arise during the sandbox experimentation, as well as response mechanisms to promptly mitigate those risks and, where necessary, stop the processing (Article 59(1), point (c)).
- Any personal data to be processed are in a functionally separate, isolated and protected data processing environment under the control of the prospective provider, and only authorised persons have access to them; the provider can further share the originally collected data only in accordance with Union data protection law, and personal data created in the sandbox cannot be shared outside the sandbox (Article 59(1), points (d) and (e)).
- The processing of personal data in the sandbox neither leads to measures or decisions affecting the data subjects nor affects the application of their rights laid down in Union law on the protection of personal data (Article 59(1), point (f)).
- The personal data processed in the sandbox are protected by appropriate technical and organisational measures and deleted once participation in the sandbox has terminated or the personal data has reached the end of its retention period, and the logs of the processing are kept for the duration of the participation in the sandbox, unless provided otherwise by Union or national law (Article 59(1), points (g) and (h)).
- A complete and detailed description of the process and rationale behind the training, testing and validation of the AI system is kept together with the testing results as part of the technical documentation referred to in Annex IV (Article 59(1), point (i)).
- A short summary of the AI project developed in the sandbox, its objectives and expected results is published on the website of the competent authorities; this obligation does not cover sensitive operational data in relation to the activities of law enforcement, border control, immigration or asylum authorities (Article 59(1), point (j)).
- For the prevention, investigation, detection or prosecution of criminal offences or the execution of criminal penalties, including safeguarding against and preventing threats to public security, under the control and responsibility of law enforcement authorities, the processing of personal data in an AI regulatory sandbox is based on a specific Union or national law and subject to the same cumulative conditions (Article 59(2)).
- This permission is without prejudice to Union or national law which excludes processing of personal data for other purposes than those explicitly mentioned in that law, and to Union or national law laying down the basis for the processing of personal data necessary for developing, testing or training innovative AI systems or any other legal basis, in compliance with Union law on the protection of personal data (Article 59(3)).
These lines are LexLint's reading of the law's text, as of .
Who enforces it
Settledness
- As of
- Open questions
- Does the substantial public interest condition of Article 59(1), point (a), reach an AI system a private company develops in the sandbox for a commercial purpose within one of the five listed areas?
What it makes you log
Log retention
Point (h) keeps the processing logs for the duration of the participation in the sandbox unless Union or national law provides otherwise; point (g) deletes the personal data once participation has terminated or the data's own retention period has ended. No fixed number of days, months or years is stated.
- As of
- Basis
- Purpose bound
- Prohibits retaining
- the personal data processed in the sandbox, once participation in the sandbox has terminated or the personal data has reached the end of its retention period
Logging duty
Point (h) makes the logs of the processing of personal data in the sandbox a condition of the permission, to be kept for the duration of the participation in the sandbox unless Union or national law provides otherwise. The point names no content for the logs; point (i) separately requires a description of the training, testing and validation process to be kept with the technical documentation.
- Kind
- Explicit
- As of
- Provision
- Article 59(1), point (h)
- Trigger
- personal_data_processing
What this law does
Personal data lawfully collected for other purposes may be processed in an AI regulatory sandbox solely for the purpose of developing, training and testing certain AI systems in the sandbox when all of the listed conditions are met.
The AI systems must be developed for safeguarding substantial public interest, by a public authority or another natural or legal person, in one or more of these areas: public safety and public health, a high level of protection and improvement of the quality of the environment, energy sustainability, safety and resilience of transport systems and mobility, critical infrastructure and networks, and efficiency and quality of public administration and public services.
The data processed must be necessary for complying with one or more of the requirements referred to in Chapter III, Section 2, where those requirements cannot effectively be fulfilled by processing anonymised, synthetic or other non-personal data.
There must be effective monitoring mechanisms to identify whether any high risks to the rights and freedoms of the data subjects may arise during the sandbox experimentation, as well as response mechanisms to promptly mitigate those risks and, where necessary, stop the processing.
Any personal data to be processed must be in a functionally separate, isolated and protected data processing environment under the control of the prospective provider with access limited to authorised persons, the provider may share the originally collected data further only in accordance with Union data protection law, and personal data created in the sandbox cannot be shared outside the sandbox.
The processing of personal data in the sandbox must neither lead to measures or decisions affecting the data subjects nor affect the application of their rights laid down in Union law on the protection of personal data.
Personal data processed in the sandbox must be protected by appropriate technical and organisational measures and deleted once participation in the sandbox has terminated or the data has reached the end of its retention period, and the logs of the processing must be kept for the duration of the participation unless Union or national law provides otherwise.
A complete and detailed description of the process and rationale behind the training, testing and validation of the AI system must be kept together with the testing results as part of the technical documentation referred to in Annex IV.
A short summary of the AI project developed in the sandbox, its objectives and expected results must be published on the website of the competent authorities, but that obligation does not cover sensitive operational data in relation to the activities of law enforcement, border control, immigration or asylum authorities.
For the prevention, investigation, detection or prosecution of criminal offences or the execution of criminal penalties, including safeguarding against and preventing threats to public security, under the control and responsibility of law enforcement authorities, the processing of personal data in an AI regulatory sandbox must be based on a specific Union or national law and subject to the same cumulative conditions as in paragraph 1.
Article 59(1) is without prejudice to Union or national law which excludes processing of personal data for other purposes than those explicitly mentioned in that law, and to Union or national law laying down the basis for the processing of personal data necessary for developing, testing or training innovative AI systems or any other legal basis, in compliance with Union law on the protection of personal data.
Article 2(7), as replaced by Regulation (EU) 2026/1744, provides that, without prejudice to Articles 4a and 59, the Regulation does not affect Regulation (EU) 2016/679 or (EU) 2018/1725, or Directive 2002/58/EC or (EU) 2016/680. Article 59 sits in Chapter VI, to which Article 113 gives no date of its own, so it applies from the Regulation's general application date of .
Regulation (EU) 2026/1744 replaced points (a) and (c) of Article 113's third paragraph and added point (d), and none of them names Chapter VI. Under Article 2(2), as replaced by Regulation (EU) 2026/1744, Articles 57, 58 and 59 apply to a high-risk AI system related to a product covered by the Union harmonisation legislation listed in Section B of Annex I only in so far as the requirements for high-risk AI systems under the Regulation have been integrated in that Union harmonisation legislation.
When LexLint raises it
When your app profile says your app makes high-risk automated decisions or trains models.