More than 100 companies have signed on as initial supporters of the EU AI Pact, a voluntary effort tied to the EU AI Act. The European Commission announced the supporter list, which includes some of the most visible companies in artificial intelligence and cloud computing.
The pact is designed to move companies toward the principles of the EU AI Act before the law takes effect. It is not only a signal of support. It asks participants to begin preparing internal practices around governance, risk identification and AI literacy.
Who signed the EU AI Pact
The list includes OpenAI, Amazon, Microsoft and Google. It also includes European businesses from several sectors, including IT, telecom, healthcare, banking and automotive.
That mix matters because AI deployment is no longer limited to companies that build foundation models or consumer chatbots. The source article describes a broader group of businesses that use, integrate or manage AI systems in different operational settings.
For companies in these sectors, signing the pact points to an early willingness to align with the EU AI Act's direction. The commitments are voluntary, but they create a public benchmark for how companies say they will approach AI governance before formal legal obligations begin.
What the voluntary commitments cover
The pact centers on three main areas. These are practical internal steps rather than product announcements or technical specifications.
- Creating an AI governance strategy: Companies commit to building a structure for how AI is managed inside the organization.
- Identifying potentially high-risk AI systems: Signatories are asked to examine where their AI systems may require closer attention under the principles of the EU AI Act.
- Promoting AI literacy among employees: The pact also emphasizes preparing staff to understand and work with AI systems responsibly.
OpenAI said the pact's focus on AI literacy, deployment and governance aligns with its mission of developing safe AI that benefits society broadly. That statement places the pact in the same general frame as many current debates around responsible AI: not just what systems can do, but how organizations oversee them.
For readers tracking AI regulation, the key point is that the pact functions as an early preparation mechanism. It asks companies to act before the law takes effect, giving the market a visible sign of which organizations are ready to engage with the EU AI Act's principles now.
Extra commitments go beyond the core pledge
The European Union says more than half of the signatories have agreed to further commitments. These go beyond the three core areas and address more specific governance concerns.
According to the source article, those additional commitments include ensuring human oversight, mitigating risks and transparently labeling certain types of AI-generated content such as deepfakes.
Those points show how the pact connects high-level AI governance to concrete user-facing concerns. Human oversight speaks to who remains accountable when AI systems are used. Risk mitigation focuses attention on possible harms before they escalate. Labeling AI-generated content, including deepfakes, addresses the problem of people not knowing whether what they see was created or altered by AI.
The pact therefore creates two layers of participation. All signatories are associated with the core voluntary commitments, while more than half have taken on additional commitments that address oversight, risk and transparency in more direct ways.
Why Apple and Meta stand out by not signing
Apple and Meta are notably missing from the supporter list. Their absence is significant because both companies are central to the consumer technology and AI landscape described in the source article.
The source article says Apple is currently not bringing AI systems into the EU, citing concerns with the DMA. Meta, meanwhile, is lobbying against EU regulations and has claimed "regulatory uncertainties."
The article also describes possible reasons for their opposition. These include not wanting to subject their AI systems to high-risk regulations, disclose their data and, in Meta's case, collect user data for AI training.
Meta has also stopped officially offering its open-source multimodal Llama models for download in the EU. That detail connects the policy dispute to product availability, showing that regulatory concerns can influence what AI systems are offered in a market and under what conditions.
What this means for AI companies in Europe
The EU AI Pact is voluntary, but it is still a meaningful signal. Over 100 companies are choosing to associate themselves with AI Act principles before those requirements formally take effect.
For companies that sign, the pact can serve as a public commitment to prepare governance systems, review potentially high-risk AI uses and improve AI literacy among employees. For companies that do not sign, especially major names like Apple and Meta, the absence becomes part of the broader discussion about how the technology industry is responding to European AI regulation.
The split also shows that AI regulation is not only a legal question. It is becoming a business strategy issue, a product availability issue and a trust issue. The companies backing the pact are presenting early alignment as part of their approach to AI deployment, while the missing names highlight the unresolved tensions around compliance, data disclosure and regulatory uncertainty.