AI Pact

What is the AI Pact?

In 2021, the Commission presented its proposal for the AI Act which is currently in the final stages of the legislative process. A political agreement was reached by the co-legislators in December 2023 and the final text should be approved in the coming weeks.

Some provisions of the AI Act will already apply shortly after the adoption of the regulation. However, some requirements on the high-risk AI systems and other provisions will only be applicable at the end of a transitional period (ie., the time between entry into force and date of applicability). 

In this context, the Commission is promoting the AI Pact, seeking the industry’s voluntary commitment to anticipate the AI Act and to start implementing its requirements ahead of the legal deadline. To gather participants, the first call for interest was launched in November 2023, obtaining responses from over 550 organisations of various sizes, sectors, and countries. The AI Office has since initiated the development of the AI Pact, which is structured around two pillars: 

• Pillar I acts as a gateway to engage the AI Pact network (those organisations that have expressed an interest in the Pact), encourages the exchange of best practices, and provides with practical information on the AI Act implementation process; 

• Pillar II encourages AI system providers and deployers to prepare early and take actions towards compliance with requirements and obligations set out in the legislation. 

Two pillars’ approach 

Pillar I: gathering and exchanging with the AI Pact network

Under Pillar I, participants contribute to the creation of a collaborative community, sharing their experiences and knowledge. This includes workshops organised by the AI Office which provide participants with a better understanding of the AI Act, their responsibilities and how to prepare for its implementation. In turn, the AI Office is can gather insights into best practices and challenges faced by the participants.

In this context, participants can share best practices and internal policies that may be of use to others in their compliance journey. Depending on participants’ preferences, these best practices may also be published online in a platform where the AI Office will share information on the AI Act’s implementation process. 

Pillar II: facilitating and communicating company pledges 

The purpose of this pillar is to provide a framework to foster the early implementation of some of the measures of the AI Act. This initiative encourages organisations to proactively disclose the processes and practices they are implementing to anticipate compliance. Specifically, companies providing or deploying AI systems can demonstrate and share their voluntary commitments towards transparency and high-risk requirements and prepare early on for their implementation. 

The commitments take the form of pledges which are ‘declarations of engagement’. These pledges contain concrete actions (planned or underway) to meet the AI Act’s distinct requirements and include a timeline for their adoption. Such declarations of engagement can also take the form of incremental objectives.

Organisations committing to preparatory actions will be asked to report on their commitments on a regular basis. The commitments are collected and published by the AI Office to ensure visibility, increase accountability and credibility, and to strengthen confidence in the technologies developed by the organisations which have committed to the pledges.

What are the benefits of the Pact for participants?

The Commission is working together with the participants and supporting them in:

• building a common understanding of the objectives of the AI Act

• taking concrete actions to understand, adapt and prepare for the future implementation of the AI Act (e.g. build internal processes, prepare staff and self-assess AI systems)

• sharing knowledge and increasing the visibility and credibility of the safeguards put in place to demonstrate trustworthy AI

• building additional trust in AI technologies

The Pact is allowing front-runners and ambitious participants to test and share their solutions with the wider community.

Tentative Timeline

Upcoming work on pledges

Organisations that have expressed interest in the AI Pact have been invited to a workshop in September to discuss the pledges, and to provide feedback on a new draft with a broader scope of commitments. The AI Office will review the feedback received and it will consider modifying the pledges as deemed appropriate. The final version of the pledges will be disclosed and discussed during the workshop, with the aim of collecting the official signatures on the pledges by the second half of September.

Interested to be an active stakeholder in the AI Pact? Join the initiative on behalf of your company!

Your company has already produced internal guidelines or processes to ensure the design, development and use of trustworthy AI systems? Share and test them with the AI Pact community!