Meta Rejects Signing EU AI Pledge, Focuses on Compliance with AI Act

Meta Platforms has decided not to immediately join the European Union's AI Pact, according to a spokesperson on Tuesday. The AI Pact serves as a temporary measure before the AI Act, a comprehensive set of rules governing artificial intelligence development and deployment, becomes fully effective.

Understanding the AI Act

The AI Act, approved by EU lawmakers in May, requires companies to provide detailed summaries of the data used to train their AI models. The majority of the Act’s rules will start applying on August 2, 2026. Until then, companies are encouraged to sign up for the voluntary AI Pact, which aims to implement key obligations of the forthcoming law.

"We welcome harmonized EU rules and are focusing on our compliance work under the AI Act at this time," a Meta spokesperson stated, hinting that the company may consider joining the AI Pact later.

The AI Pact’s Role and Expectations

The AI Pact is designed to bridge the gap until the AI Act is fully operational. It encourages companies to proactively engage with the new regulations and fosters collaboration to develop best practices for AI governance. Signatories are expected to commit to three core actions:

• Adopting an AI Governance Strategy: Fostering AI adoption within the organization and preparing for compliance with the AI Act.

• Mapping High-Risk AI Systems: Identifying and assessing AI systems that could be categorized as high-risk under the AI Act.

• Promoting AI Literacy: Ensuring ethical and responsible AI development by enhancing awareness and education among staff.

Voluntary Pledges and Flexibility

The AI Pact offers a list of potential pledges, allowing signatories to choose commitments that align with their strategies. These pledges include designing AI systems to inform users when interacting with AI, and clearly labeling AI-generated content, such as deep fakes.

This flexible approach aims to encourage competition among companies to showcase their commitment to AI safety and responsibility. The European Commission’s list of over 100 signatories includes notable companies like Amazon, Microsoft, OpenAI, Samsung, Salesforce, Qualcomm, Palantir, and SAP. However, some major players, such as Apple and Meta, are notably absent.

Meta’s Compliance Strategy

Meta’s decision to focus on compliance with the AI Act rather than joining the AI Pact may be influenced by the stringent penalties for non-compliance. Fines can reach up to 7% of global annual revenue for severe violations; up to 3% for failing to meet other AI Act requirements, and up to 1.5% for providing inaccurate information to regulators, making it critical for the company to prioritize adherence to the law’s requirements.

Notable Absences and Criticisms

Other notable absentees include Anthropic, Nvidia, and Spotify. Spotify’s absence is particularly surprising given its recent involvement in lobbying against regulations that could hinder generative AI’s growth. Meanwhile, German AI company Aleph Alpha has joined the Pact but is shifting its focus to providing B2B support for generative AI tools, potentially signaling a change in its policy priorities.

Implications of the AI Pact

The AI Pact is a proactive step by the EU to engage companies in responsible AI development ahead of the AI Act’s full implementation. While the initial list of signatories represents a broad range of industries, the voluntary nature of the pledges means their impact remains to be seen. Companies are invited to report on their progress 12 months after publishing their pledges, which could lead to another wave of public scrutiny and recognition.

For a complete list of early AI Pact signatories, visit the European Commission's official website.