(Updated  Sept 1. 2025)

Dear Valued Customer,

As you intend to use and/or integrate our Bria 3.2 model, we want to ensure you have all the information needed to successfully incorporate our AI model into your systems. This Model Documentation Form is part of our commitment to transparency and compliance related to our development of General-Purpose AI models, in compliance with some specific provisions of the EU AI Act, which came into effect in August 2025.

Why are we sharing this information?

As a downstream provider integrating Bria's general-purpose AI model into your systems, you need comprehensive technical details to:

• Understand our model's capabilities
• Meet your own compliance obligations under the AI Act
• Build trust with your end users through transparency

What is included as part of this form?

This documentation provides detailed information about our model's technical specifications, training methodology, licensing terms, and acceptable use policies. We've organized the information to help you quickly find what matters most for your specific use case, whether you're building creative tools, enhancing workflows, or developing new AI-powered products.

Our Commitment to Partnership

Bria has signed the EU's General-Purpose AI Code of Practice, demonstrating our dedication to responsible AI development. This Form includes all the information to be documented as part of Measure 1.1 of the Transparency Chapter of the Code of Practice. By providing this detailed documentation proactively, we're not just meeting regulatory requirements, but we are supporting your success and ensuring our partnership is built on transparency and trust.

If you have questions about any aspect of this documentation or need additional technical details for your integration, please don't hesitate to reach out to our team. We're here to support your AI journey every step of the way.

Best regards,

Bria Team

Public Notice: Bria AI Compliance with EU AI Act "GPAI Code of Practice" Requirements

We are pleased to announce that Bria has successfully complied with the Code of Practice Requirements under the EU AI Act and is now making available its Copyright Policy and Public Summary of Training Content as required by law. We have also gone beyond our legal responsibilities and published the comprehensive Model Documentation Form - a detailed technical specification document that we are only required to provide to our downstream customers upon request, but which we are now making publicly available to demonstrate our unwavering commitment to transparency and industry leadership.

At Bria, we believe that innovation and responsibility go hand in hand. As a leading generative AI company building rights-cleared, attribution-based models for enterprise use, we are proud to be among the first to demonstrate full compliance with the European Union's groundbreaking AI Act requirements.

What did Bria implement:

✓ Transparency Commitment: We have successfully implemented the transparency commitments of the EU's General-Purpose AI Code of Practice, demonstrating our dedication to responsible AI development and transparent business practices.

✓ Public Documentation Available: Our comprehensive Public Summary of Training Content is now publicly accessible, providing detailed information about our model training data, sources, and compliance measures.

✓ Comprehensive Copyright Policy: Bria has implemented and published a comprehensive Copyright Policy that establishes our framework for copyright compliance in AI development, including exclusive use of commercially licensed training data, proprietary attribution technology for fair compensation to data licensors, robust technical safeguards to prevent copyright-infringing outputs, rigorous due diligence procedures for third-party data sources, and accessible communication channels for rightsholder concerns.

Access Our Compliance Documentation:

You can now access our Copyright Policy, Public Summary of Training Content, and detailed Model Documentation Form, which provide comprehensive information about:

• Training data sources
• Technical specifications and model capabilities
• Copyright compliance policies and procedures
• Data processing methodologies and safeguards

Our Commitment to Responsible AI:

This compliance milestone reflects our core belief that generative AI can be both innovative and ethical. Through our proprietary attribution technology and exclusive use of licensed training data, we continue to demonstrate that responsible AI development is not only possible but commercially viable.

We invite you to explore our compliance documentation and discover how Bria is setting new standards for trustworthy AI development in the European market and beyond.

For questions about our AI Act compliance or to access our documentation, please get in touch with us at legal@bria.ai.

Effective Date: September 1^st, 2025 

Public Summary of Training Content for General-Purpose AI models required by Article 53 (1)(d) of Regulation (EU) 2024/1689 (AI Act)

s

2.       List of data sources

2.1.    Publicly available datasets

Fully licensed images provided by Bria data partners through commercial licensing agreements with rightsholders globally. All training data is sourced exclusively through transactional commercial licensing agreements that explicitly authorize the use of licensed content for generative AI model training purposes. Each licensing agreement includes comprehensive warranties and representations from data licensors confirming their full legal right, title, and authority to license the data objects and grant the rights necessary for AI training applications.

2.2. Private non-publicly available datasets obtained from third parties

2.2.1 Datasets commercially licensed by rightsholders or their representatives

2.2.2 Private datasets obtained from other third parties

2.3. Data crawled and scraped from online sources

2.4. User data

2.5. Synthetic data

According to the AI Office's instructions, this Section does not refer to the use of AI models to clean or enrich data (e.g., AI-generated metadata to enrich or modify a dataset, such as creating text descriptions of images). Bria does use captions generated with third-party AI models for enrichment of its data.

2.6.    Other sources of data

3. Data processing aspects

• Respect for the reservation of rights from the text and data mining exception or limitation

3.2.    Removal of illegal content

General description of measures taken: Bria’s training data sourcing methodology inherently prevents illegal content inclusion through exclusive use of commercially licensed datasets obtained from verified professional data partners and established image bank providers. All training data consists of human-created content with explicit commercial use releases, deliberately excluding public figures, harmful materials, copyrighted fictional characters, or any unlawful content. Our comprehensive due diligence procedures require all data providers to warrant their full legal right and authority to license content, provide representations confirming non-infringement of third-party intellectual property rights, and maintain detailed chain-of-title documentation. We implement content moderation filters to prevent the generation of harmful or biased content and enforce strict guidelines against generating harmful or offensive content. Our data collection process emphasizes cultural sensitivity and maintains rigorous permissions systems with periodic compliance audits to ensure continued adherence to legal standards. Since Bria does not engage in web-crawling or utilize publicly accessible online content, the risk of incorporating illegal or unauthorized material is systematically eliminated at the source.

Bria Copyright Policy

(updated September 2025)

1. Copyright Compliance Framework at Bria:

Our commitment to copyright compliance is fundamental to our responsible AI development. We recognize that true accountability in AI requires rigorous adherence to intellectual property rights and legal standards.

Our copyright policy reflects our core principles: respecting rightsholders, ensuring full legal compliance, maintaining the highest standards of ethical AI development, and providing equitable compensation to all data licensors through our proprietary and patented algorithmic attribution technology that measures and rewards each contributor’s impact on generated content.

This document outlines our comprehensive approach to copyright compliance for our general-purpose AI models placed on the market.

We have developed this policy to demonstrate our commitment to:

• Fully complying with laws and regulations on copyright and related rights, specifically Article 53(1)(c) of the EU AI Act, and Article 4(3) of Directive (EU) 2019/790 on copyright and related rights in the Digital Single Market.
• Identifying and respecting the rightsholder's reservations of rights.
• Utilizing state-of-the-art technologies for copyright protection - ensuring all of our models as trained only on fully licensed training data.
• Being committed to advancing AI technology while adhering to principles that prioritize societal benefit and accountable development.

Our approach goes beyond mere legal compliance. It represents an intrinsic commitment to integrity, transparency, and responsible innovation in our AI models development and deployment processes.

By successfully building high-quality models through sustainable data partnerships, Bria demonstrates that responsible innovation and respect for intellectual property are not only possible but also commercially viable. This evidence is crucial for policymakers, regulators, and courts, showing there is no need to choose between fostering AI progress and protecting creators.

2.  Training Data Compliance and Verification

• Licensed Data Use Policy: Bria exclusively utilizes commercially licensed training data explicitly authorized for generative AI model training. We maintain comprehensive documentation and records of all of our licensing agreements and authorizations provided to us by our data partners.
• Data Source Verification: Our data licensing process includes:
  • Legal verification of all licensing agreements confirming that:
    • Bria has entered into valid and enforceable agreements with all data licensors that expressly authorize the use of licensed data for generative AI model training purposes, and that
    • Each licensing agreement includes comprehensive warranties and representations from data licensors confirming their full legal right, title, and authority to license the data objects and grant the rights necessary for AI training applications.
    • Maintenance of detailed data cataloging systems;
    • Regular audits of data source compliance.
• No Web-Crawling Policy: Bria does not and will not engage in web-crawling activities or utilize publicly accessible online content for model training. This approach ensures complete compliance with copyright obligations and eliminates risks associated with unlicensed content.
  • Rights Reservation Compliance: While our business model does not involve web-crawling, we acknowledge and respect rightsholder reservations expressed pursuant to Article 4(3) of Directive (EU) 2019/790 and maintain systems to identify and comply with such reservations, if applicable.

3.  Data Partner Compensation and Attribution Framework

• Fair Compensation: We compensate partners based on their measured contribution to generated content, providing a recurring revenue stream to data partners. Bria rejects one-time license deals for AI training data, instead implementing a fair and transparent system of ongoing compensation that ensures creators receive a share of the long-term value their work generates in existing and future AI applications.
• Attribution Technology: Our patented attribution technology enables Bria to bridge the gap between demand for synthetic content and supply of authentic training data. Each synthetic output generated by our models is attributed to the original content creators that impacted the most on its generation.
• Transparency: Bria’s attribution technology offers transparency to both data contributors and customers.

4. Content Due Diligence for Third-Party Sources:

• Bria maintains comprehensive due diligence procedures for any content obtained through third-party providers. We require all third-party data providers to:
  • Warrant their full legal right and authority to license content for AI training purposes;
  • Provide comprehensive warranties that all data licensed is original and human-created content;
  • Ensure all data is subject to legal releases covering commercial use and AI training applications where identifiable persons or property are depicted;
  • Provide representations confirming non-infringement of third-party copyrights, trademarks, or other intellectual property rights;
  • Provide warranties that content does not feature public figures, harmful materials, or copyrighted fictional characters;
  • Present all applicable compliance certifications regarding adherence to all applicable copyright laws and regulations;
  • Maintain detailed records and chain-of-title documentation, and
  • Be subject to periodic compliance audits and remediation procedures.
• Data Procurement and Usage. In addition, Bria maintains the following processes to ensure all data licensed and used by it will not subsequently create any other harmful implications on rightsholders or users:
  • Data Quality and Diversity: No synthetic data is used for training our models. Bria does not acquire data that will not lead to model quality improvement and will only dilute the attribution to existing partners.
  • Dedicated Storage and Access: Training data for generally available models is stored in dedicated training accounts. Proprietary data provided by customers for training private models is segregated from general models’ training data accounts. Access to proprietary data provided by customers for training private models and models informed by customers’ proprietary data is restricted to such customers and authorized Bria support for maintenance functions only with rigorous permissions system. Access to models is subject to periodical penetration tests.
  • Asset Cataloging: An automated pipeline records each asset in a dedicated catalog, maintaining a clear link to its origin.

5. Output Protection and Copyright Compliance

Bria implements robust technical and policy measures to prevent copyright-infringing outputs from our general-purpose AI models:

• Technical Safeguards. We employ state-of-the-art technical measures to avoid our models from reproducing training content in an infringing manner, including:
  • Advanced content filtering systems;
  • Real-time output monitoring;
  • Regular testing protocols to verify safeguard effectiveness;
  • Detection of potential copyright violations post-generation.
• Acceptable Use Requirements. Our terms and conditions explicitly prohibit:
  • Using models to generate copyright-infringing content;
  • Circumventing technical safeguards;
  • Using our models to create any unlawful content.
• Documentation Requirements. For all model deployments, we provide:
  • Clear copyright and acceptable-use compliance guidelines;
  • Explicit prohibition of infringing uses; and
  • Technical documentation of protective measures.
• Cross-Platform Application. These protective measures apply consistently across direct model implementations, on-prem model integrations, API access and all other deployment scenarios.
• Monitoring and Enforcement. We maintain ongoing oversight through:
  • Regular audits of model outputs;
  • Compliance verification procedures;
  • Incident response protocols; and
  • Continuous improvement of safeguards.

6. Rightsholder Contact and Complaint System.

We maintain accessible communication channels and procedures for rightsholders through an email message to legal@bria.ai, to submit precise and substantiated complaints regarding non-compliance with copyright commitments, potential infringement concerns, rights reservation violations or any other copyright-related matters.

This reporting mechanism complements but does not replace or limit any legal remedies available under Union and national copyright law.

7. Other protective measures

Beyond our comprehensive copyright compliance framework, Bria maintains additional protective measures to ensure full regulatory adherence and responsible deployment practices for the benefit of both our data licensors and end users. These supplementary safeguards reflect our commitment to establishing industry-leading standards that transcend basic legal requirements and encompass the full spectrum of ethical AI development considerations.

• Privacy and Data Protection
  • We place strong emphasis on privacy protection for individuals. We use human likenesses only with explicit releases for commercial use. We strictly prohibit the use of public figures' data in our models.
  • We implement robust data protection measures to safeguard personal information and comply with relevant privacy and data protection regulations.
• Identity Protection and Training Architecture
  • No Identity Mapping. Bria’s training datasets are fully anonymized through the removal of all naming and identifying metadata. Our models are designed to prevent any mechanism for linking visual likenesses to real person identities.
  • Balanced, Multi-Source Training Catalog. We maintain a wide-ranging, globally sourced dataset that deliberately balances input across creators, contexts, content types, and personalities. This diversity reduces prominence of any single source and minimizes potential for overrepresentation or unintended memorization.
  • Architecture Design. Bria’s models are specifically architected to learn general visual patterns rather than memorize or reproduce specific individuals. Our technical architecture and sampling methodology inherently discourage identity persistence or overfitting to any single visual subject, making it highly improbable for even specific prompts to generate recognizable individuals.
• Commitment to Diversity, Inclusion, and Bias Mitigation
  • We actively seek and incorporate diverse datasets that represent a wide range of cultures, ethnicities, ages, genders, and geographical locations. This approach ensures our AI models are trained on a comprehensive representation of global diversity.
  • Our development process includes rigorous testing and refinement to identify and mitigate potential biases in our AI models. We employ diverse teams of experts to continuously evaluate and improve our models' fairness and inclusiveness.
  • We implement content moderation filters to prevent the generation of harmful or biased content.
  • We regularly consult with diverse stakeholder groups, including underrepresented communities, to understand their perspectives and incorporate their feedback into our AI development and deployment processes.
  • Our AI models are designed with cultural sensitivity in mind, respecting and accurately representing various cultural nuances and contexts in their outputs.
  • We regularly publish reports on our diversity and inclusion initiatives, including the composition of our datasets and the results of our bias mitigation efforts, to maintain accountability and encourage industry-wide progress.
• Safety, Mitigation of Misinformation and Harmful Content
  • All AI-generated content is distinctly marked using cutting-edge technologies.
  • We enforce strict guidelines against generating harmful or offensive content.
  • We implement measures to prevent the misuse of our platform. We do not accept any data that features public figures or harmful content, making it impractical to use our models to generate these concepts.

To conclude:

This Copyright Policy establishes Bria’s comprehensive framework for responsible AI development through five core pillars: (1) exclusive use of commercially licensed training data with verified legal authorization, (2) proprietary attribution technology that provides ongoing compensation to data licensors based on measured contribution impact, (3) robust technical safeguards preventing copyright-infringing outputs, (4) rigorous due diligence procedures for third-party data sources, and (5) accessible rightsholder communication channels for compliance concerns.

Our approach encompasses additional protective measures including privacy protection, diversity and inclusion initiatives, bias mitigation protocols, and safety measures against harmful content generation.

Through our data partner empowerment and collaboration framework, we maintain long-term partnerships that create mutual value while advancing responsible AI innovation.

Many of our data licensors leverage our models to offer their creator communities and customers advanced generative AI capabilities under preferred commercial arrangements. This symbiotic relationship strengthens our collaborative ecosystem and delivers mutual value creation across the AI development lifecycle.

We view our data licensors as integral to our sustainable AI development approach and remain committed to nurturing long-term, mutually beneficial relationships that evolve with the advancing AI landscape and regulatory environment.

Together, we can create an AI landscape that is not just innovative but also trustworthy, inclusive, and truly beneficial for all.