In a bid to ensure a smooth and proactive transition to the upcoming AI regulations, the European Commission has introduced the AI Pact, a voluntary initiative designed to help organizations prepare for the full implementation of the AI Act, which came into force on August 1, 2024. Although some provisions are already applicable, the requirements for high-risk AI systems will only come into effect after a transitional period. The AI Pact serves as a framework to encourage early compliance and set the stage for responsible AI development and deployment.
What is the AI Pact?
The AI Pact is an initiative created to support organizations in aligning with the requirements of the AI Act ahead of the mandatory deadlines. It promotes voluntary participation from companies across various sectors, encouraging them to begin implementing measures related to high-risk AI systems. In its initial phase, launched in November 2023, the AI Pact attracted interest from over 550 organizations worldwide. The Pact is structured around two key pillars:
- Pillar I: Focuses on building a collaborative network where participants can exchange best practices and gain insights into the AI Act implementation process.
- Pillar II: Encourages companies to take early steps toward compliance by making voluntary pledges to implement AI governance strategies and adhere to high-risk AI system requirements.
Pillar I: Collaborative Exchange and Knowledge Sharing
Under Pillar I, participants join a network where they can share experiences and knowledge about implementing AI regulations. The AI Office hosts workshops and collaborative forums to help organizations understand their responsibilities under the AI Act and prepare for compliance. This engagement also allows participants to share internal best practices that can be beneficial to others navigating similar challenges.
By sharing this knowledge, participants help build a community that fosters innovation while maintaining compliance with AI safety standards. The AI Office will make these insights available on an online platform, providing a valuable resource for others looking to adopt similar strategies.
Pillar II: Early Compliance Through Voluntary Pledges
Pillar II focuses on providing a framework for organizations to commit to early compliance with the AI Act. Companies are encouraged to make voluntary pledges outlining the steps they will take to align with the Act’s high-risk system requirements. These pledges are designed to promote transparency and accountability, allowing companies to showcase their proactive efforts in the development and deployment of trustworthy AI.
The voluntary pledges may include a range of actions, such as:
- Implementing AI governance strategies to promote responsible AI development.
- Identifying and mapping AI systems that may fall under the high-risk category in the AI Act.
- Promoting AI literacy and awareness among employees to ensure the ethical use of AI technologies.
Participants are also encouraged to tailor their commitments to their specific activities, such as incorporating human oversight of AI systems, mitigating potential risks, and transparently labeling AI-generated content, particularly deepfakes.
The Role of the Pledges
On September 25, 2024, over 100 companies from sectors including IT, healthcare, banking, and automotive industries signed the first round of pledges. These voluntary commitments, though not legally binding, allow companies to take a leadership role in AI regulation. The pledges offer a structured way for companies to demonstrate their commitment to AI safety and prepare for the full applicability of the AI Act.
Participants are expected to publicly report their progress 12 months after signing the pledges, offering transparency and accountability. This reporting process will help both participants and regulators monitor the progress of AI safety measures across industries.
Benefits of Joining the AI Pact
The AI Pact offers several advantages for participating organizations. By joining, companies can:
- Gain a deeper understanding of the objectives of the AI Act and prepare for its implementation.
- Develop internal processes to ensure compliance, such as creating governance structures and training staff on AI safety protocols.
- Share knowledge and increase the visibility of their efforts to ensure AI is trustworthy and safe.
- Build trust in AI technologies, enhancing the credibility of their AI systems both internally and with the public.
Additionally, the AI Pact allows participants to test new solutions and share their findings with the wider community, positioning them as front-runners in AI innovation and safety.
