Data protection
Global Privacy Assembly Adopts Resolutions on Agentic AI
Over 130 data protection authorities adopted resolutions establishing privacy principles for autonomous AI agents and artificial intelligence in education.
On November 20, 2026, the French Data Protection Authority (CNIL) (opens in a new tab) published the outcomes of the 48th Annual Meeting of the Global Privacy Assembly. The assembly took place in Brussels, Belgium, from November 16 to 18, 2026. The Global Privacy Assembly brings together more than 130 data protection authorities from over 80 countries. During this meeting, the authorities adopted five resolutions, including two focused on artificial intelligence.
The assembly adopted the Resolution on Privacy Protective Agentic Artificial Intelligence to address the challenges and opportunities presented by AI agents. These systems have the ability to process personal data and make automated decisions. The resolution recalls that existing personal data protection obligations apply to these systems. It also highlights the importance of human oversight when deploying agentic artificial intelligence.
The assembly also adopted the Resolution on Artificial Intelligence in Education (opens in a new tab). This text addresses the protection of children, noting that they are particularly exposed to risks related to digital uses. The resolution encourages the education ecosystem to consider the challenges of artificial intelligence, given the sensitive nature of collecting and using personal data in schools.
The education resolution outlines 11 core principles to safeguard personal data. The primary principle is the best interests of the child, which must guide every stage of data processing. Organizations must identify roles and responsibilities beforehand. They must determine a legal basis and strictly respect the intended purpose of the processing. The principles include conducting data protection impact assessments and ensuring privacy by design and by default.
The remaining principles for educational artificial intelligence focus on data accuracy and limiting retention periods. The resolution calls for the regulation of automated decision making and the implementation of measures to guarantee data security. Organizations must ensure transparency and accountability. They must also establish data processing agreements and procurement rules. Finally, the principles require guarantees regarding the rights of children, legal guardians, and education professionals.
The Global Privacy Assembly included six commitments in the education resolution. The authorities call for strategic international cooperation to establish common standards for artificial intelligence governance in education. The commitments include informing and raising awareness about privacy risks in schools. The assembly encourages authorities to use the resolution to develop complementary guidelines and promote it to policymakers. The authorities also committed to liaising with international organizations, insisting on the need for a legal framework for artificial intelligence in educational settings, and sharing resources and best practices.
Beyond artificial intelligence, the assembly adopted resolutions on privacy enhancing technologies, data sharing for combating money laundering, and the establishment of a practice hub. The network also identified wearable devices, with a focus on smart glasses, and privacy preserving age assurance as priority topics for knowledge sharing in 2027.
Who it applies to
The resolutions set expectations for organizations developing or deploying artificial intelligence systems that process personal data. The resolution on agentic artificial intelligence applies to technology providers and organizations using autonomous agents that make automated decisions.
The resolution on artificial intelligence in education applies to the broader education ecosystem. This includes technology providers, educators, public authorities, parents, and students. Any entity involved in the processing of personal data through artificial intelligence in educational environments falls under the scope of these principles. The Global Privacy Assembly represents data protection authorities across 80 countries, making these principles relevant across multiple jurisdictions.
What it requires
Organizations deploying agentic artificial intelligence must apply personal data protection obligations to their autonomous systems. The resolution requires organizations to maintain human oversight over these systems.
For artificial intelligence in education, organizations must prioritize the best interests of the child at every processing stage. They must identify roles, establish a legal basis, and conduct data protection impact assessments. Technology providers must ensure privacy by design, maintain data accuracy, and limit how long they retain personal data. Organizations must secure the data, ensure transparency, and provide mechanisms for children and their guardians to exercise their data protection rights. They must also regulate automated decision making and formalize data processing agreements.
The dates
The Global Privacy Assembly adopted the resolutions during its 48th annual meeting in Brussels, held from November 16 to 18, 2026. The CNIL published the official announcement detailing the texts on November 20, 2026. The next annual meeting of the Global Privacy Assembly will take place in Nairobi, Kenya, in 2027.
What to check now
- Identify any autonomous artificial intelligence agents currently deployed that process personal data or make automated decisions.
- Verify that human oversight mechanisms are in place for systems using agentic artificial intelligence.
- Review the legal basis and purpose for any artificial intelligence tools used in educational settings.
- Confirm that data protection impact assessments are completed for artificial intelligence systems processing student data.
- Check data retention policies to ensure educational data is not kept longer than necessary.
- Ensure that contracts and data processing agreements with educational technology providers align with privacy requirements.
The resolutions from the Global Privacy Assembly indicate a coordinated approach among international data protection authorities. As artificial intelligence systems gain autonomy and enter sensitive environments like schools, regulators expect organizations to apply existing privacy frameworks rigorously and maintain strict human oversight.
Sources
- CNIL (opens in a new tab) · www.cnil.fr
- CNIL (opens in a new tab) · www.cnil.fr
This article is general information, not legal advice. For your company's situation, talk to a qualified lawyer.
This article was written with AI and checked, sentence by sentence, against the sources linked above.