GLTaC & AI

AI Regulation Effort in the U.S.

While there isn't a single, comprehensive federal law governing AI in the US, there's a growing body of regulations and initiatives at both the federal and state levels. Here's a breakdown of the key aspects:

Federal Level:

• Executive Order 14110 (October 2023) on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence: This executive order directs federal agencies to develop AI safety and security, protect privacy and civil liberties, address algorithmic discrimination, and promote responsible innovation. It directs various federal agencies to develop guidelines and regulations related to AI.
• National Artificial Intelligence Initiative Act of 2020 (NAII): This act, while primarily focused on promoting AI research and development, also directs the National Institute of Standards and Technology (NIST) to develop standards for AI systems.
• NIST AI Risk Management Framework: Voluntary guidelines for responsible AI development.
• Federal Trade Commission (FTC): The FTC has been active in enforcing existing laws, such as those related to consumer protection, in the context of AI. They have taken action against companies for deceptive or unfair practices involving AI.
• EEOC: Providing guidance addressing AI discrimination in employment.
• FDA: Focusing on oversight of AI medical devices.
• Consumer Financial Protection Bureau (CFPB): Exploring how existing laws apply to AI and may issue further guidance or regulations.

State Level:

• Colorado AI Act: AI legislation requiring regular testing and risk management. This is one of the most comprehensive state-level AI laws, focusing on regulating the use of AI in insurance.
• California: Requires businesses to assess the impact of automated decision tools.
• New York City Local Law 144: Regulates AI in employment decisions.
• Other State Laws: Several other states have passed or are considering legislation related to specific aspects of AI, such as its use in hiring, facial recognition, and automated decision-making.

Industry-Specific:

• Financial services: Banking regulators are examining AI use in lending and risk assessment.
• Healthcare: FDA regulation of AI medical devices through the Software as a Medical Device framework.

AI Regulation Beyond the U.S.

AI regulation is a hot topic worldwide, and different countries are taking varied approaches. Here's a glimpse at some key regions and their strategies: 

Europe:

• European Union (EU): The EU is leading the charge with its AI Act (March 2024). The world's first comprehensive AI law, which is a landmark piece of legislation using a risk-based approach. It classifies AI systems based on risk levels and imposes corresponding obligations. This act is expected to set a global precedent for AI regulation. This takes effect in phases between 2024 and 2026. The various risk levels include the following:
• Banned: Unacceptable risk AI systems (social scoring, emotion recognition in workplaces)
• High-risk systems: Requires extensive documentation, testing, and oversight
• General purpose AI: Subject to transparency requirements

• United Kingdom (UK): The UK is taking a more innovation-friendly approach, focusing on principles and guidance rather than hard law. They are emphasizing collaboration with industry and international cooperation on AI safety. The key factors are as follows:
• Pro-innovation approach with voluntary principles
• AI Safety Institute established in 2023
• Regulatory framework focused on existing regulators applying AI principles

Asia:

• China: Generative AI regulations (August 2023) require content moderation and data protection. China has a comprehensive AI regulatory regime, focusing on algorithmic systems with respect to transparency and preventing algorithmic discrimination, deep synthesis information services, and generative AI. China has sectoral regulations for autonomous vehicles and financial services. Their regulations emphasize upholding Chinese values and state stability.
• Singapore: Singapore is a leader in AI governance, with a Model AI Governance Framework and a focus on sector-specific oversight and ethical testing requirements. They have also been quick to adapt their regulations to new developments like generative AI.
• Japan: Japan has a "human-centered" approach to AI, emphasizing ethical principles and social responsibility. They are currently taking a "soft law" approach, with guidelines rather than strict regulations, pursuing voluntary guidelines and international cooperation.

Other Regions:

• Canada: Canada is developing its own AI strategy, with the Artificial Intelligence & Data Act (AIDA) currently under consideration, which would regulate high-impact AI systems. It includes an Algorithmic Impact Assessment tool for government AI use. They are aiming to balance innovation with responsible AI adoption.
• Australia: Australia is taking a similar approach to Canada, with proposed guardrails for high-risk AI and a voluntary standard for AI safety.
• Brazil: Currently has a draft legislation based on EU model under consideration.
• India: Is working on a risk-based approach similar to EU AI Act.