The European Union has officially enforced new transparency and safety regulations for general-purpose AI systems, starting August 2, 2025. These rules apply to companies developing large language models, generative tools, and AI systems used across multiple sectors. This marks a significant step in the EU’s broader AI Act agenda, setting a global precedent for regulating AI technologies.
Under the new rules, companies like OpenAI, Google, Meta, and Anthropic must provide documentation on training datasets, system performance, risk management, and compliance procedures. Transparency obligations require disclosures about AI-generated content, potential biases, and decision-making processes.
Failure to comply may result in hefty penalties—up to 7% of a company’s global revenue. While some tech firms have already adapted by publishing model cards and governance reports, others are scrambling to meet the EU’s high standards.
These new rules are part of the EU’s effort to ensure that AI systems used by millions of citizens uphold values like human rights, privacy, and accountability. The regulation also introduces “high-risk” categories—such as AI used in healthcare, education, and employment—which are subject to stricter rules and assessments.
Tech firms have expressed mixed responses. While some applaud the clarity, others warn of slowed innovation and fragmented global compliance requirements. Nonetheless, this enforcement could become a blueprint for other countries seeking to regulate AI without stifling innovation. The global AI landscape may now shift towards greater transparency and ethical oversight.