AI Legislation 2024: US & Global Updates

Here's what you need to know about AI laws in 2024:

• EU AI Act takes effect August 1, 2024 - first major global AI regulation
• US lacks comprehensive federal AI law, but states are acting
• China, Japan, and other countries developing their own approaches

Key focus areas:

• Privacy and data protection
• Transparency and explainability
• Fairness and non-discrimination
• Safety and security
• Intellectual property rights

What this means for businesses:

• Start preparing for compliance now
• Assess AI systems for risk levels
• Expect more regulations globally

Quick Comparison of Major AI Regulations:

Region	Key Regulation	Approach	Enforcement
EU	AI Act	Risk-based tiers	Fines up to €35M or 7% of revenue
US	No federal law yet	Sector-specific rules	Varies by industry/state
China	Multiple targeted rules	Focus on content and algorithms	Strict with domestic enforcement

The AI legal landscape is evolving rapidly. Stay informed and start preparing your AI governance now to adapt to coming changes.

Related video from YouTube

Global AI Rules Overview

The world's racing to regulate AI. Here's who's leading the charge:

Joint International Efforts

Several big players are setting the stage for global AI governance:

1. OECD

• Adopted AI ethics principles in 2019
• G20 leaders backed these principles
• Now a hub for AI policy harmonization

2. UNESCO

• All 193 member states adopted AI ethics recommendations in 2021
• A big step towards a global ethical AI framework

3. G7 and the Hiroshima AI Process

• Launched in 2023
• Aims to boost AI governance cooperation

4. Council of Europe

• Working on legally binding AI rules
• Draft published in December 2023

5. Global Partnership on AI (GPAI)

• Started in 2020 by 15 countries
• Supports ethical AI adoption worldwide

But here's the catch: global AI governance is still a mess. There's not enough teamwork between institutions, creating a weak 'regime complex'.

The EU's leading the pack with its AI Act - the world's first AI law. It's likely to set the tone for regulations elsewhere.

The US is taking baby steps. Colorado's the first state to regulate AI, with rules for high-risk systems starting in 2026. The feds issued an AI executive order in late 2023, focusing on risk management.

Japan's jumping in too, considering rules to make big AI developers spill the beans on certain info.

As AI keeps evolving at breakneck speed, these global efforts will shape its future. The trick? Balancing innovation and safety while navigating the maze of international politics and tech advances.

US AI Policy Changes

The US is shaking up its AI policy game. Both federal and state governments are making moves.

Federal Updates

The Federal Artificial Intelligence Risk Management Act of 2024 is a big deal:

• Federal agencies must use NIST's AI Risk Management Framework
• OMB has 180 days to issue guidance after NIST guidelines
• NIST gets one year to provide agency guidance

Plus, it's bringing AI experts into federal agencies and developing risk management guidelines.

State Laws

States aren't waiting for Congress:

Colorado is leading the pack:

• First state with comprehensive AI rules
• Signed May 17, 2024
• Focuses on consumer protection
• Kicks in February 1, 2026

Utah and Tennessee are close behind:

• Utah: "Artificial Intelligence Amendments" (April 2024) - AI interaction disclosure
• Tennessee: "ELVIS Act" (March 2024) - Fights AI-generated voices and fake recordings

Many states are setting up AI task forces or study groups.

Some are zeroing in on government AI use:

State	Action
Maryland	Government AI safeguards
New Hampshire	Government AI safeguards
Virginia	Banned certain high-risk government AI uses

Key Trends:

• States crafting their own AI approach
• More focus on commercial and private sector AI rules
• Growing need for comprehensive regulations

For businesses, this means:

• Build strong AI governance programs
• Document AI design, development, and deployment
• Do regular risk and impact assessments

Expect more state-level AI laws to pop up fast in 2024.

EU AI Act Explained

The EU AI Act, approved in March 2024, is the first major AI regulation worldwide. It's designed to protect EU citizens while encouraging AI innovation.

Here's how it works:

The Act splits AI systems into risk categories:

Risk Level	What It Means	Examples
Unacceptable	Banned	Social scoring, AI manipulation
High	Strict rules	Biometric ID, education assessment
Limited	Must be transparent	Chatbots, deepfakes
Minimal/Low	No special rules	AI in video games

Following the Rules

You need to comply if you:

• Sell AI systems in the EU
• Serve EU users with AI
• Use AI outputs in the EU

To comply:

1. List your AI: Know what AI you're using or selling.

2. Check the risk: Figure out which category your AI falls into.

3. Meet the requirements: For high-risk AI, you'll need to:

• Manage risks
• Keep detailed tech docs
• Ensure humans are involved
• Register in the EU database

4. Get ready: You have 36 months from publication to fully comply.

"The AI Act will be enforced EU-wide and by national authorities in each EU country."

Breaking the rules is expensive:

Violation	Max Fine
Using banned AI	€35 million or 7% of global turnover
High-risk AI rule breaks	€15 million or 3% of global turnover
Lying to authorities	€7.5 million or 1% of global turnover

The EU AI Act is setting the bar for AI rules globally. If you want to do business in the EU, start getting ready now.

AI Rules in Asia

China, Japan, and Singapore are leading the AI regulation charge in Asia. Each takes a unique approach to balance innovation and risk.

China: Targeted and Active

China's not messing around. They've rolled out specific AI rules:

Regulation	Date	What It Does
Generative AI Measures	Aug 15, 2023	Labels AI content, applies to all services in China
Deep Synthesis Provisions	Jan 10, 2023	Tackles deep fakes, bans illegal info
Ethics Review Measures	Dec 1, 2023	Sets ethical standards for AI development

China's focusing on specific AI uses, not one big law. The Cyber Administration of China (CAC) now wants companies to file their public opinion-influencing algorithms.

What this means for businesses? Get your docs in order and be ready to file. As of June 30, 2024, over 1,400 AI algorithms from 450+ companies have been filed in China.

Japan: Cautious and Watching

Japan's taking it slow:

• No specific AI law yet
• Using existing rules and guidelines
• Thinking about a basic AI law

Japan's economy ministry says binding AI rules aren't needed now. But that could change as AI evolves.

Singapore: Light Touch and Voluntary

Singapore's keeping it flexible:

• No AI-specific rules
• Voluntary frameworks for ethics and governance
• Using existing sector rules for AI risks

They're aiming to boost innovation while guiding responsible AI development.

What It Means for You

1. Stay Sharp: AI rules in Asia are changing fast. Keep an eye out, especially if you're in multiple countries.

2. Get Ready: Even without big laws, be prepared. China's algorithm filing is just the start.

3. Ethics Matter: All three countries care about ethical AI. Make it a priority in your projects.

4. Be Flexible: Be ready to tweak your AI strategies for different Asian markets.

As AI grows, expect more changes across Asia. Stay flexible and proactive with compliance and ethical AI development.

sbb-itb-ea3f94f

AI Plans in Other Regions

African Union's AI Approach

The African Union (AU) is working on a continent-wide AI policy. Their draft aims to:

• Create Africa-specific AI rules
• Set up industry standards
• Establish testing grounds and national AI councils

Seven African countries already have their own AI strategies. The AU's plan won't be final until February 2025.

Some key points:

• AI could add $136 billion to four major African economies by 2030
• Experts debate whether to grow AI first or regulate it
• Real-world uses include:
  • Tanzanian farmers spotting crop diseases
  • South African researchers studying housing issues
  • Kenyan authorities analyzing security footage

South Africa's AI Framework

South Africa is leading with its National AI Policy Framework:

Focus	Goal
Talent	Build AI skills
Tech	Improve digital systems
Ethics	Create AI guidelines
Safety	Ensure secure AI

The plan puts humans first, aiming to enhance, not replace, human decisions. It should be ready by September 2024.

South Africa's AI sector is booming:

• Worth $0.90 billion now (2024)
• Could hit $4.00 billion by 2030
• Microsoft just invested $70 million (May 2024)

Middle East AI Projects

The UAE and Saudi Arabia are shaping AI policies:

Country	Plans
UAE	AI Strategy 2031, Ethics Guidelines, Generative AI Rules
Saudi Arabia	National Data & AI Strategy, AI Authority (SDAIA)

The UAE wants an AI-friendly system by 2031. Saudi Arabia aims to lead in AI by 2030.

Other Gulf countries are joining in:

• Bahrain: Launched an AI lab
• Qatar: Created a national AI plan

Egypt introduced AI ethics rules in April 2023, following OECD principles.

While there's no unified Middle East AI plan, countries are making progress. They're balancing new tech with ethics by:

1. Boosting AI investment
2. Protecting privacy and ideas
3. Matching global AI standards

Expect more updates as AI keeps changing.

Main Topics in 2024 AI Laws

The AI legal landscape in 2024 is all about balancing innovation with protection. Here's what's hot:

Privacy and Data Protection

AI loves data, but privacy comes first. New laws focus on:

• Tighter data rules
• Labeling AI-generated content
• Giving users control over their data

The EU AI Act, kicking in August 1, 2024, makes high-risk AI developers show their homework on data use and testing.

Transparency and Explainability

No more black boxes. Laws want AI to be open:

What's Required	Why It Matters
AI disclosure	You know when you're talking to a bot
Explaining decisions	Understanding why AI said "no"
Documenting models	Showing how the AI sausage is made

The U.S. Algorithmic Accountability Act aims to make big AI users explain their impact.

Fairness and Non-discrimination

AI bias is a no-go. Laws are cracking down on:

• Unfair AI in hiring and lending
• Biased outcomes in AI systems
• Lack of diversity in AI teams

The EEOC's warning: AI in hiring can discriminate. Watch out!

Safety and Security

AI safety is getting serious:

• New safety institutes popping up globally
• Risk checks for high-impact AI
• Battling AI-powered fraud and cyber threats

The FTC's worried about AI voice scams. It's getting real out there.

Intellectual Property Rights

Who owns AI stuff? It's complicated:

• AI-generated content ownership debates
• New rules for AI training data
• Can AI be an inventor? The jury's still out

The EU AI Act says GenAI providers must play nice with copyright laws.

Consumer Protection

Keeping consumers safe from AI shenanigans:

• No AI trickery in marketing
• Fighting AI manipulation on social media
• Right to challenge AI decisions

Utah's new law holds AI users responsible for consumer law violations.

"There is no AI exemption to the laws on the books." - Lina Khan, FTC Chair

Translation: AI isn't above the law.

As AI grows, expect these themes to keep shaping the rules, trying to keep up with the tech while protecting society.

How AI Laws Affect Business

AI laws are changing how companies work. Here's what's happening:

Industry Impact

Healthcare and finance are seeing big shifts:

Industry	Key Changes
Healthcare	- Tighter privacy rules for AI diagnostics - More testing for AI medical tools
Finance	- AI lending decisions under watch - AI trading needs more openness

In healthcare, AI medical devices now need proof they're safe. IDx-DR, an AI that spots eye problems in diabetics, had to pass tough tests first.

Banks are changing too. JPMorgan Chase uses AI to catch fraud but must explain how it works.

Global Rule Challenges

Companies face different AI laws worldwide:

• EU rules affect anyone serving EU customers
• China limits AI using Chinese data
• U.S. has a mix of state laws

IBM, for example, follows EU rules for Watson in Europe and California rules in the U.S.

Dasha Simons from IBM says: "The AI Act has different rules for different roles. You need to know which apply to you."

To handle this:

1. Know your AI: Track what AI you use and where.

2. Check risks: Sort AI systems by how much harm they could cause.

3. Build expert teams: Mix legal and tech know-how.

4. Use smart tools: AI can help manage global rules.

5. Keep learning: Laws change fast. Stay informed.

Breaking rules is expensive. EU fines can hit €35 million or 7% of global sales.

As laws change, businesses must adapt. It's about following rules AND building trust in AI.

What's Next for AI Laws

The AI legal landscape is changing fast. Here's what's coming:

More US Federal Action

The US is playing catch-up with the EU on AI laws. In 2023, Congress saw over 30 AI bills. Now, there's pressure to act.

Senator Chuck Schumer's SAFE Innovation Framework aims to balance AI growth and risk:

Principle	Goal
Security	Guard against AI threats
Accountability	Make AI creators responsible
Foundations	Back AI research
Explainability	Clarify AI decisions
Innovation	Keep US competitive

EU AI Act Takes Effect

The EU AI Act starts in August 2024. It'll shape global AI rules by sorting AI into risk levels:

• Unacceptable risk: Banned
• High risk: Strict rules
• Limited risk: Some oversight

Breaking the rules? Fines up to €35 million or 7% of global sales.

Global Coordination Challenges

Countries have different AI goals. The US-China tech rivalry makes global rules tough.

AI experts Mark Nitzberg and John Zysman say:

"The key question is how to create a system where tools work across borders while respecting different preferences."

New AI Watchdogs

Tech leaders want new AI regulators. OpenAI's Sam Altman suggests an agency to license big AI projects and check safety.

But ex-Google chair Eric Schmidt warns:

"No one in government can get AI oversight right."

The challenge? Making rules that work for fast-changing tech.

Focus on AI Safety and Ethics

AI safety is a top concern. Expect more talks on:

• AI bias and fairness
• Data privacy
• AI's job impact

Industry Self-Regulation

While waiting for laws, companies might set their own rules. The EU's AI Pact invites firms to follow the AI Act early.

Legal Industry Shifts

AI is changing law work. Deloitte thinks 114,000 legal jobs could go. But new roles will pop up too.

Legal tech expert Richard Susskind stresses:

"Without transparency and explainability, we can't understand AI decisions. Without XAI, there's no trust."

What This Means for You

1. Stay informed: AI laws will affect many industries.

2. Plan ahead: Start thinking about AI compliance now.

3. Be ready to adapt: Laws will likely change as AI grows.

The next few years are key for AI laws. They'll shape how we use and govern AI for years to come.

Conclusion

AI laws are changing fast in 2024. Here's what you need to know:

EU's Big Move: The EU AI Act kicks in August 1, 2024. It's a game-changer:

Risk Level	What Happens
Unacceptable	Not allowed
High	Tough rules
Limited	Some checks

Break the rules? You could face fines up to €35 million or 7% of global sales.

US Catching Up: No big AI law yet, but things are moving:

• Congress is talking about the Algorithmic Accountability Act
• California and New York are pushing their own AI rules
• Federal agencies are paying more attention to AI

Global AI Rules: Countries are making their own AI laws:

• EU: Big, risk-based plan
• US: Rules for specific sectors
• China: Cares about data and national security

What This Means for You:

1. More rules are coming. Start getting ready now.

2. Figure out your risk level under the EU AI Act.

3. EU rules might become the norm everywhere.

What to Do in 2024:

• Check your AI systems
• Set up AI rules in your company
• Keep an eye on new laws

Sam Altman from OpenAI says:

"The key question is how to create a system where tools work across borders while respecting different preferences."

AI laws are tricky and always changing. Stay informed to handle these changes well.

FAQs

What's the deal with AI laws in the US?

There's a new bill on the block: SB970. Introduced in January 2024, it's shaking things up:

• It wants AI content tool sellers to give users a heads-up about legal risks
• Users need to know they could face charges if they misuse the tech

This bill shows how the US is trying to tackle AI risks without killing innovation.

What AI laws exist right now?

The US doesn't have a big, all-encompassing AI law. Instead, we've got:

• A White House Executive Order on AI (it's more like guidelines)
• Some proposed bills at federal and state levels
• Rules for specific industries that sometimes apply to AI

Level	AI Rules
Federal	No big law, just orders and guidelines
State	A few states have some AI laws
Industry	Some sectors have their own AI rules

Are other countries regulating AI?

You bet. Here's what's happening:

• EU: Their AI Act kicks in August 2024 - it's a world-first
• China: They're all about data and national security
• UK: Working on their own AI rules

The AI law scene is changing fast. The US even started an AI safety institute to check out risks from advanced AI models.

Erik Brynjolfsson from Stanford says: "The folks in Brussels, they come up with a lot of bureaucratic rules that make it harder for companies to innovate."

It's a tough balance: we want innovation, but we also need some rules of the road.

Related posts

• AI Compliance Guide: International Trade Law 2024
• AI Compliance Guide for Legal Tech Startups 2024
• AI Disclosure Laws 2024: Business Requirements
• AI-Powered Regulatory Change Monitoring Tools 2024