AI Regulations New
A sourced reference on AI Regulations.
What is the EU AI Act?
The EU AI Act is the world's first comprehensive legal framework for artificial intelligence, adopted by the European Parliament in March 2024. It classifies AI systems by risk level—unacceptable, high, limited, and minimal—and imposes obligations on developers and deployers accordingly. [Source: European Parliament]
When does the EU AI Act take effect?
The EU AI Act entered into force on August 1, 2024, with a phased implementation. Prohibitions on unacceptable-risk AI apply from February 2025, rules for general-purpose AI models from August 2025, and obligations for high-risk systems fully apply by August 2026. [Source: European Commission]
What counts as 'high-risk' AI under the EU AI Act?
High-risk AI systems under the EU AI Act include those used in critical infrastructure, education, employment, essential public services, law enforcement, migration, and administration of justice. These systems must meet strict requirements for transparency, human oversight, accuracy, and technical documentation before deployment. [Source: European Commission]
What AI practices are banned under the EU AI Act?
The EU AI Act prohibits AI systems that use subliminal manipulation, exploit vulnerable groups, enable social scoring by governments, perform real-time remote biometric identification in public spaces (with limited exceptions), and conduct untargeted scraping of facial images for recognition databases. [Source: European Parliament]
What are the fines for violating the EU AI Act?
Violations of the EU AI Act carry fines of up to €35 million or 7% of global annual turnover for prohibited AI practices, €15 million or 3% for other violations, and €7.5 million or 1.5% for providing incorrect information to regulators—whichever is higher. [Source: European Commission]
What did the US Executive Order on AI require?
President Biden's Executive Order 14110, signed October 30, 2023, directed federal agencies to develop AI safety standards, required developers of powerful AI models to share safety test results with the government, and instructed agencies to address AI risks in areas including national security, privacy, and civil rights. [Source: The White House]
What is Trump's executive order on AI from 2025?
On January 20, 2025, President Trump signed Executive Order 14179, 'Removing Barriers to American Leadership in Artificial Intelligence,' which revoked Biden's 2023 AI order and directed agencies to develop a new AI Action Plan prioritizing innovation and removing regulations deemed burdensome to AI development. [Source: The White House]
Has the United States passed a federal AI law?
As of mid-2025, the United States has not enacted a comprehensive federal AI law. Regulation has proceeded through executive orders and sector-specific agency guidance. The EU AI Act remains the only major jurisdiction with a standalone AI regulatory statute, though U.S. states like California have advanced their own legislation. [Source: Congressional Research Service]
What is the NIST AI Risk Management Framework?
NIST's AI Risk Management Framework (AI RMF 1.0), published in January 2023, is a voluntary guidance document helping organizations identify, assess, and manage risks posed by AI systems across four core functions: Govern, Map, Measure, and Manage. It is widely referenced in U.S. federal AI policy. [Source: NIST]
What is the US AI Safety Institute?
The U.S. AI Safety Institute (AISI), established within NIST under the 2023 Executive Order, conducts research and develops guidelines for evaluating the safety of advanced AI models. In 2025, the Trump administration rebranded it as the AI Safety and Security Board, refocusing its mandate. [Source: NIST]
What is the UK AI Safety Institute?
The UK AI Safety Institute (AISI), launched in November 2023 following the Bletchley Park AI Safety Summit, conducts advanced safety research on frontier AI models and tests systems for dangerous capabilities. It published its first evaluations framework in 2024 and operates under the Department for Science, Innovation and Technology. [Source: UK Government]
What was the Bletchley Park AI Safety Summit?
The Bletchley Park AI Safety Summit, held on November 1–2, 2023, was the first major intergovernmental gathering on frontier AI safety. Twenty-eight countries, including the US, UK, China, and EU members, signed the Bletchley Declaration committing to cooperative action on AI risks and establishing national AI safety institutes. [Source: UK Government]
How does the EU AI Act regulate general-purpose AI models like GPT?
The EU AI Act places tiered obligations on general-purpose AI (GPAI) models. All GPAI providers must maintain technical documentation and comply with copyright law. Providers of models with systemic risk—those trained with over 10^25 FLOPs—face additional duties including adversarial testing, incident reporting, and cybersecurity obligations. [Source: European Commission]
Which US states have passed AI legislation?
As of 2025, over 40 U.S. states have introduced AI-related legislation. Colorado's SB 205 (2024) covers high-risk AI in consequential decisions. Illinois, Texas, and Utah have AI laws targeting employment. California's legislature passed multiple AI bills in 2024, though Governor Newsom vetoed the broad SB 1047 safety bill. [Source: National Conference of State Legislatures]
Are there regulations on using AI in hiring decisions?
New York City Local Law 144 (effective July 2023) requires employers using AI hiring tools to conduct annual bias audits and disclose AI use to candidates. The EEOC has also issued guidance stating that employers remain liable under Title VII for discriminatory outcomes from AI screening tools. [Source: NYC Department of Consumer and Worker Protection]
What is algorithmic accountability in AI regulation?
Algorithmic accountability refers to obligations requiring organizations to assess, explain, and mitigate harms caused by automated decision-making systems. The FTC has exercised enforcement authority under Section 5 of the FTC Act against deceptive or unfair AI practices, and has issued guidance specifically on AI bias and transparency. [Source: Federal Trade Commission]
What do AI transparency requirements mean in practice?
AI transparency requirements oblige deployers to disclose when users are interacting with AI systems, explain automated decisions that significantly affect individuals, and maintain documentation of model design and training data. The EU AI Act and the NIST AI RMF both include detailed transparency provisions applicable to different risk tiers. [Source: NIST]
Does the EU AI Act apply to US companies?
Yes. The EU AI Act has extraterritorial reach: it applies to any provider or deployer whose AI system outputs are used in the EU, regardless of where the company is based. U.S. companies deploying AI services to EU residents or businesses must comply with its risk-tier obligations. [Source: European Commission]
How does China regulate artificial intelligence?
China regulates AI through a series of targeted rules: the 2022 Deep Synthesis Regulation (deepfakes), the 2022 Recommendation Algorithm Regulation, and the 2023 Generative AI Measures, which require security assessments, content labeling, and user real-name registration for publicly deployed generative AI services. [Source: Cyberspace Administration of China]
What are the OECD AI Principles?
The OECD AI Principles, adopted in May 2019 and updated in 2024, are the first intergovernmental standard on AI, endorsed by over 46 countries including the G20. They establish five value-based principles—inclusive growth, human-centered values, transparency, robustness, and accountability—as a foundation for national AI policy. [Source: OECD]