Google has announced its endorsement of the Code of Practice on Transparency of AI-Generated Content, formulated within the framework of the EU AI Act. The code outlines four key requirements: the use of machine-readable markings and watermarks, transparent disclosure of human-AI interactions, mandatory labeling of deepfake content, and the identification of texts relating to public interests. This initiative is in harmony with Google's 2025 Global Partnership on AI Code of Conduct and reinforces its commitment to advancing the C2PA content provenance standard and developing AI transparency tools, such as SynthID.

Google remains dedicated to promoting SynthID digital watermarking technology and is actively collaborating with various third-party AI organizations to encourage the industry-wide adoption of interoperable watermarking solutions. At the same time, Google has voiced concerns regarding the escalating complexity of regulatory frameworks. The company notes that overly intricate regulations may clash with Europe’s objectives of enhancing competitiveness and simplifying processes. Additionally, such complexity could potentially confuse protected groups, impeding their ability to access straightforward and comprehensible background information.