The European Union on Sunday expanded the scope of its Artificial Intelligence Act, bringing into force new transparency requirements and rules governing providers of general-purpose artificial intelligence (AI) models.

The new rules, which took effect on August 2, require interactive AI systems such as chatbots to clearly inform users that they are interacting with AI rather than a human.

Under the transparency requirements, AI-generated or AI-manipulated content, including images, videos and audio known as "deepfakes," must be clearly labeled. AI-generated or modified content must also include machine-readable markers to make it easier to identify, according to the European Commission.

The Commission said the measures are intended to reduce deception and manipulation associated with AI-generated content.

The rules also empower the EU AI Office to enforce provisions of the AI Act on providers of general-purpose AI models.

Providers of advanced general-purpose AI models that may pose systemic risks are subject to additional obligations aimed at mitigating large-scale harms, including risks related to cyberattacks, loss of model control, harmful manipulation and violations of fundamental rights.

The EU's Artificial Intelligence Act entered into force on August 1, 2024, establishing a risk-based regulatory framework that classifies AI systems into four categories: unacceptable risk, high risk, transparency risk, and minimal or no risk.