EU AI Act: What changes as new transparency rules take effect from August 2

Beginning August 2, the European Union will implement new transparency requirements for content created or modified using artificial intelligence. Under Article 50 of the EU Artificial Intelligence Act, selected AI-generated materials—including deepfakes and AI-produced content related to public interest topics—must carry clear labels indicating their AI origin.

To support organizations in meeting these requirements, the European Commission has released a voluntary Code of Practice that outlines recommended transparency measures for AI providers and deployers.

Since the AI Act applies to any AI systems or services offered within the EU market, its influence is expected to extend beyond European borders. Major technology companies such as Meta and Google have already endorsed the voluntary code, demonstrating their commitment to following the new guidelines.

Understanding the EU AI Act

Adopted in 2024, the EU AI Act is the world's first comprehensive legislation dedicated to regulating artificial intelligence. It follows a risk-based framework, where regulatory obligations increase according to the potential impact an AI system may have on health, safety, or fundamental rights.

Under this approach:

• AI systems classified as presenting unacceptable risk are prohibited.
• High-risk AI applications must comply with extensive governance, documentation, and oversight requirements.
• Certain AI technologies, including many generative AI models, must satisfy transparency obligations.
• Low-risk AI applications face minimal regulatory requirements.

The transparency measures introduced under Article 50 are part of the Act's phased implementation and aim to ensure users can easily recognize AI-generated content and identify when they are interacting with AI-powered systems.

Content Covered by the Transparency Rules

The European Commission identifies three primary categories that require disclosure:

• AI-generated or manipulated deepfakes.
• AI-generated text addressing matters of public interest.
• AI systems that interact directly with users, such as chatbots.

The required disclosures must be visible, easy to understand, and clearly presented. The Commission also recommends standardized labels such as "AI," "AI-Generated," and "AI-Modified" to promote consistency across digital platforms.

Exemptions and Enforcement

The regulation provides limited exemptions in specific circumstances. Creative works—including artistic, satirical, fictional, or entertainment content—may follow a lighter disclosure approach, provided the notice does not interfere with the audience's experience.

Similarly, AI-generated content that has undergone substantial human editorial review and is published under the responsibility of a legally accountable individual or organization may not require additional disclosure.

Certain law enforcement activities involving AI are also exempt when the technology is used for preventing, detecting, investigating, or prosecuting criminal offenses.

Organizations that fail to comply with Article 50 could face administrative penalties of up to €15 million or 3% of their annual global turnover, whichever amount is greater.

EU vs. India: A Different Regulatory Approach

Unlike the European Union, India has not yet introduced a dedicated AI law. Instead, AI-generated content is primarily governed through amendments to the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021.

Recent updates by India's Ministry of Electronics and Information Technology require users and social media platforms to clearly identify synthetically generated content.

Although both regions promote transparency, their regulatory focus differs. The EU places greater responsibility on AI developers and service providers by embedding transparency requirements directly into AI systems while also defining exemptions for creative works and human-reviewed content.

India's framework places more responsibility on users and digital platforms. Individuals are expected to disclose AI-generated content, while online platforms are responsible for monitoring compliance and removing content that violates existing regulations.

Voice Of Osiz

The EU AI Act marks a significant step toward building a more transparent and accountable AI ecosystem. As global regulations evolve, businesses must prioritize responsible AI practices alongside innovation. Embedding transparency, compliance, and governance into AI solutions will not only meet regulatory expectations but also strengthen user trust. At Osiz, we believe that future-ready AI solutions should be designed with ethics, security, and compliance at their core. Organizations that proactively align with emerging AI regulations will be better positioned to scale confidently across global markets. Responsible AI is no longer just a regulatory requirement—it is becoming a key driver of sustainable digital transformation.

Source: www.business-standard.com