OpenAI Supports EU Code of Practice for Transparency in AI-Generated Content

Artificial intelligence is transforming how people create, edit, and share content online. In this context, one question becomes increasingly urgent: how can we tell if what we see online was created by a human or a machine? This is precisely the issue the European Commission seeks to address with its new Code of Practice on Transparency in AI-Generated Content — and OpenAI has just announced its formal support for the initiative.

What is the European Commission's Code of Practice?

The Code of Practice on Transparency in AI-Generated Content is a regulatory initiative from the European Commission that represents a concrete step in implementing the EU AI Act, the European legislation on artificial intelligence considered one of the most comprehensive in the world. The central goal of the code is to ensure that citizens, platforms, and institutions can identify and contextualize content produced or modified by AI systems.

The code was developed with the participation of hundreds of stakeholders — tech companies, civil society organizations, regulatory bodies, and academic experts — and establishes clear guidelines for digital ecosystem actors to adopt provenance standards in their products and services.

OpenAI's Position

In announcing its support for the code, OpenAI emphasized that the European initiative aligns with work the company has been developing internally for years. According to the company, the commitment to transparency in AI-generated content is not new: since 2024, OpenAI began incorporating C2PA (Coalition for Content Provenance and Authenticity) metadata into its image generation tool DALL-E 3.

Since then, the company has made significant progress in this area:

• Improvements in labeling and detection methods for AI-generated content;
• Development of new research on digital provenance;
• Launch of a public verification tool, allowing anyone to check the origin of specific content.

Why Does Digital Provenance Matter?

Digital provenance — the trail of information about the origin and history of a piece of content — is one of the fundamental pillars for combating misinformation in the era of generative AI. With increasingly powerful tools capable of generating images, videos, audios, and extremely realistic texts, the ability to verify the authenticity of information becomes essential for democratic health.

Among the main benefits of provenance technologies are:

• Detection of large-scale misinformation campaigns;
• Protection of electoral integrity, preventing public opinion manipulation with false content;
• Empowerment of the end-user, who gains more context to interpret what they consume online;
• Accountability of platforms and creators of artificially generated content.

OpenAI explicitly mentioned the role of provenance in supporting electoral integrity, a sensitive topic ahead of elections in various democracies around the world in the coming years.

The C2PA Standard: A Global Alliance for Authenticity

A central element in OpenAI's strategy — and in the European Code of Practice itself — is the C2PA (Coalition for Content Provenance and Authenticity), a coalition that brings together technology companies, media outlets, and organizations to create open technical standards for tracking the origin of digital content.

The C2PA functions as a sort of "digital birth certificate" for image, video, audio, and document files. Through cryptographically signed metadata, it is possible to know who created the file, when, with which tool, and whether it has been modified after its original creation.

The participation of major players like OpenAI in the C2PA is crucial for the standard to gain traction and be adopted globally. The more companies and platforms that implement the standard, the more robust and trustworthy it becomes.

OpenAI and AI Governance in Europe

Support for the new code is not an isolated event. In 2025, OpenAI became the first American company to sign the EU General AI Use Code of Practice (GPAI Code of Practice), demonstrating a proactive stance towards European regulation — in contrast to some other large tech companies that resisted certain regulatory demands from the bloc.

For OpenAI, clear and applicable rules are not an obstacle to innovation but rather a condition for the responsible and sustainable development of AI. The company has consistently reinforced the message that regulation and innovation are not antonyms.

A Collective Effort, Not an Isolated One

One point that deserves emphasis in OpenAI's stance is the recognition that no single company can solve the issue of transparency in AI-generated content. It is a challenge that requires collaboration across the entire value chain: model developers, distribution platforms, content creators, media outlets, and, of course, users themselves.

The European Code of Practice acknowledges this reality by involving actors from different sectors in its formulation and implementation. OpenAI has committed to fulfilling the applicable requirements for its relevant products, contributing to making the European — and global — digital ecosystem safer and more trustworthy.

What to Expect Moving Forward?

With the advancement of European regulations and the growing engagement of major tech companies, the coming months are expected to bring:

• Expansion of provenance verification tools to more OpenAI products beyond DALL-E 3;
• Increased regulatory pressure for other platforms and developers to adopt similar standards;
• Development of new research on content labeling methods, including more resilient watermarking technologies against manipulation;
• International collaboration to align provenance standards across different jurisdictions, avoiding regulatory fragmentation.

Conclusion

OpenAI's support for the European Commission's Code of Practice is an important signal that the AI industry is, at least in part, moving towards greater responsibility and transparency. Digital provenance is not a magic solution to all problems associated with synthetic content, but it represents a concrete and technically feasible step towards restoring people's ability to trust — with justification — in what they see and hear online.

In a world where deepfakes, fake news, and media manipulation are real threats to democracy and social cohesion, initiatives like this are not just welcome: they are urgent and necessary.