Karsten Wenzlaff, Advisor
August 26th, 2025
Karsten Wenzlaff is a Crowdfunding and Fintech Expert from Germany. He is Secretary-General of Digit[...]
EU AI Transparency Rules Take Effect Setting New Benchmark
AI Regulation | Oct 6, 2025
Europe’s New AI Framework Raises the Bar For Transparency, Trust, and Safety
The European Union’s transparency and copyright rules for general purpose AI models formally took effect on August 2, 2025. Enforcement faced resistance from industry and calls for delay but regulators seem intent on sticking with the original timelines.
See: Founder Salaries: Insights from Europe, U.S., and Canada
Global tech leaders such as, OpenAI, Microsoft, Google, Anthropic, and Amazon have signed the EU Code of Practice for General Purpose AI Models, a framework quickly emerging as the new benchmark for responsible AI and a blueprint for innovators in Canada’s fintech and technology ecosystem.
Key Takeaways
- The EU’s new AI rules are in force. Developers must share brief summaries of their training data and follow copyright law
- The Code of Practice focuses on openness, legal data use, and safe system design
- Companies must record data sources, manage risk, and prevent copyright misuse
- Models trained with over 10²³ FLOP are treated as high-impact systems
- OpenAI, Microsoft, Google, Anthropic, Amazon, IBM, and Mistral AI are among the first signatories (Meta however hasn't signed)
- Transparency anchors trust and global competitiveness.
EU Transparency Rules Now in Force
The EU AI Act, published as Regulation (EU) 2024/1689 in the Official Journal of the European Union, moved into its second phase of implementation this summer. The new rules apply to general purpose AI models (GPAI) trained on large-scale data that can generate text, images, and audio. These models must now meet defined transparency and documentation requirements to operate in the European market.
See: Canada’s Opportunity In Efficient Reasoning AI
In July 2025, the European Commission released the Code of Practice for General Purpose AI Models, available on its AI transparency portal. Although its voluntary, the Code itself translates the Act’s broad principles into operational commitments that technology providers can follow, and is becoming a global reference for responsible AI development.
Transparency, Copyright, and Safety in Practice
The Code asks developers to be clear about how their AI systems are built. They must publish short summaries of training data and explain the steps taken to reduce bias and misinformation. It marks the first time transparency in AI has been defined as a regulated standard rather than a voluntary goal.
It also sets new expectations for how copyrighted material is handled. Companies must confirm that training data was obtained legally and that rights holders who chose to opt out were respected. They are encouraged to build safeguards that stop AI models from reproducing protected content and to keep a public channel open for complaints.
Finally, the Code highlights safety and security as core responsibilities. Models that use very large amounts of computing power must undergo risk assessments, maintain strong cybersecurity protections, and keep records of any misuse or technical incidents. This brings AI oversight closer to the risk management standards already used in finance.
Who's Committed to the Code?
The European Commission’s list of signatories includes OpenAI, Microsoft, Google, Anthropic, Amazon, IBM, and Mistral AI. Their participation gives the voluntary Code practical weight and sends a message that transparency and accountability are compatible with innovation.
See: Europe’s Digital Euro and Bank Stablecoin Race
Meta has not signed, citing uncertainty about enforcement and consistency across European member states. It shows the divide between companies moving early to align with transparency and those waiting for further legal clarification. Early signatories are setting examples that others may have to follow once regulation tightens.
The Business Case for Transparency
Companies that publish their data summaries and risk documentation not only comply with regulation but also strengthen their position with investors and partners. For companies that can responsibly demonstrate how they legally sourced their data to train their models, it can influence purchasing decisions in regulated sectors like finance, health, and insurance.
For AI developers, transparent practices can also lead to improved internal knowledge. Knowing exactly what data was used to train a system and where risks exist improves quality and helps teams innovate with confidence. Transparency is evolving from a compliance task into a type of competitive strategy.
What Canadian Innovators Can Take Away
Canada’s Artificial Intelligence and Data Act (AIDA) was part of Bill C-27, which died on the Order Paper when Parliament was prorogued in January 2025. For now, Canada's Voluntary Code of Conduct on the Responsible Development and Management of Advanced Generative AI Systems is still Canada's main AI policy reference. The Code is endorsed by companies such as Cohere, OpenText, BlackBerry, and Telus, and encourages transparency and human oversight in the use of generative AI. While not legally binding, it mirrors much of the EU’s direction and provides a practical foundation for organizations preparing for future legislation.
See: Europe’s Fintech Reset at Point Zero Forum
Canadian firms that sell or integrate AI systems in the European market must already comply with the EU AI Act transparency and copyright rules. The regulation applies outside the EU as well, meaning any company offering AI services or products in Europe must meet the same standards as European providers. This includes publishing training data summaries, documenting data sources, and ensuring lawful use of copyrighted material. For fintechs and technology companies, aligning early with EU standards is not just about compliance. It can help position them as credible and trusted partners in markets that increasingly link transparency with competitiveness.
Outlook
AI transparency and innovation are now hand in hand in the EU. While enforcement of the new framework will happen gradually, the market impact is immediate with companies now competing not only on performance but on integrity. Companies building transparency into product design, investor communications, and risk management will benefit by transforming responsible AI from an obligation to a strength.
The National Crowdfunding & Fintech Association (NCFA Canada) is a financial innovation ecosystem that provides education, market intelligence, industry stewardship, networking and funding opportunities and services to thousands of community members and works closely with industry, government, partners and affiliates to create aa vibrant and innovative fintech and funding industry in Canada. Decentralized and distributed, NCFA is engaged with global stakeholders and helps incubate projects and investment in fintech, alternative finance, crowdfunding, peer-to-peer finance, payments, digital assets and tokens, artificial intelligence, blockchain, cryptocurrency, regtech, and insurtech sectors. Join Canada's Fintech & Funding Community today FREE! Or become a contributing member and get perks. For more information, please visit: www.ncfacanada.org
 |  |  |
|---|---|---|
 |  |  |
Leave a Reply