The regulation of artificial intelligence (AI) is a pressing concern globally, with various proposals being considered to ensure its safe and responsible development. One area of focus is the regulation of deepfakes, which are AI-generated audio or video that can be used to deceive people. Some proposals suggest prohibiting deepfakes in political campaign advertisements to prevent their use in depicting opponents saying something they didn't or events that didn't occur.
Another proposal is to prohibit the public distribution of pornographic deepfakes made without the consent of the person depicted. Additionally, some lawmakers suggest labeling deepfakes shared publicly to inform viewers that the content is AI-generated.
Regulating AI decision-making is also a key area of concern. Some proposals suggest requiring AI programs to pass a test before deployment to evaluate potential biases or security vulnerabilities. Others propose allowing the government to audit AI programs in use and requiring fixes for any problems found.
International regulation of AI is also being explored. Some proposals suggest establishing an international agency to regulate large-scale AI projects and develop international standards. Others propose creating a treaty to prohibit the development of lethal autonomous weapons that can fire on targets without human control.
A risk-based approach to AI regulation is also being considered, categorizing AI systems into low, high, and unacceptable risk categories. This approach would impose stricter regulations on high-risk AI systems, such as those used in healthcare or transportation.
To ensure compliance with AI regulations, some proposals suggest imposing significant fines for non-compliance, such as up to 7% of worldwide annual turnover for launching prohibited AI systems. Establishing regulatory bodies to oversee AI development and enforcement is also being considered.
Ultimately, effective regulation of AI will require a balanced approach that promotes innovation while protecting people from potential harms. By working together, governments, industry leaders, and civil society can create a regulatory framework that supports the responsible development of AI
