File0001

Integrate cryptographic hashing and blockchain technology to create immutable records of the original content**. These tools should enable the verification of content integrity, ensuring that any alterations are detected and traced back to the source. 6.3.3 Apply visible, tamper-resistant watermarks (e.g., steganography) to all deepfake content to ensure its synthetic nature is identifiable across platforms*. In creative contexts where synthetic media is disclosed (e.g., in production credits), watermarks may be omitted. Ensure watermarks are durable across formats. 6.3.4 Implement robust version control systems for all documentation, ensuring all changes are tracked, recorded, and easily traceable. (AI Ethics Principles – Principle 7) . (AI Ethics Principles – Principle 7) (AI Ethics Principles – Principle 5; Anti-Cyber Crime Law Article 3) (AI Ethics Principles – Principle 7) *Content creators should disclose the use of deepfake technology: Provide disclaimers in the content’s description, credits, or accompanying materials, explain­ ing that deepfake technology was used. These disclosures should be visible and accessible, ensuring audiences are aware of the synthetic nature of the content -Labelling content directly, using on-screen indicators or text at the start or end of the content, to inform viewers that synthetic media was used. **Additionally, content creators could use metadata to label deepfake content, specifying the original creator and any subsequent modifications. This ensures clear attribution and responsibility for each version of the content. Further, content creators are encouraged to maintain an ownership log that records each instance of shared or remixed content, documenting any modifications and identifying responsible parties. This log aids in tracking the evolution of content and establishes accountability. (AI Ethics Principles – Principle 6) 6.4 Accountability and Responsibility Deepfakes Guidelines: Mitigating risks while fostering innovation Effective regulation of deepfake technology is crucial to safeguard against privacy and security risks. Regulators can enforce a range of measures to minimize risks, however, regulation should also focus on unlocking the benefits of deepfakes. This section aims to ensure that deepfakes are regulated while enabling potential benefits 7. Guidance for Regulators 7.1.1 Platforms should implement mechanisms for reporting harmful deepfake content and review flagged material promptly based on its risk. High-risk cases, such as finance, politics, or impersonation, should be prioritized, while low-risk or educational content may allow for longer review periods. 7.1.2 Regularly monitor these platforms to ensure compliance with directives and impose penalties for failure to act on identified harmful content. 7.2.1 Conduct thorough assessments of the deepfake technologies being used or approved for use within the country. This includes reviewing the algorithms, data sources, and ethical considerations in the devel­ opment of such technologies. 7.2.2 Develop a formal approval process for deepfake technologies, ensuring they meet established ethi­ cal, legal, and technical standards before being deployed or used commercially. 7.2.3 Develop specific risk assessment frameworks for government entities that will use or approve deep­ fake technologies. 7.2.4 Develop and enforce set standards for producing generative content, such as those outlined by the Coalition for Content Provenance and Authenticity (C2PA), to ensure transparency and accountability in the creation and dissemination of deepfake and other AI-generated media. 7.2 Risk Assessment and Approval Processes 7.1 Platform Regulation and Monitoring Deepfakes Guidelines: Mitigating risks while fostering innovation