Survey Reveals: 79% of Americans Want Strict AI Laws

The rapid proliferation of AI has brought about new and concerning dangers. In an attempt to mitigate the harmful aspects of AI, governments worldwide are responding with various laws and regulations. 

To understand the public’s concerns and expectations about governmental AI regulation, we surveyed 2,000 US residents and performed an in-depth analysis of the regulatory landscape across 195 countries. 

By examining the effectiveness of current measures and gauging the desire for stricter regulations, our research seeks to provide valuable insights that can inform policymakers, industry leaders, and the public as we navigate this uncharted territory.

Why We Conducted This Study

We performed this research to understand two things:

• How governments are progressing with AI regulations
• How the public wants governments to react to the growing perils of AI

This survey is in response to the growing risks of AI. 

Some perils are already occurring – such as using AI to scam people, create inappropriate deepfakes, and spread disinformation – while other potential threats lay on the horizon, including widespread job loss, autonomous weapons, and uncontrollable self-aware AI.   

As a result, calls for governmental regulation of AI have come from many sectors – including leaders in AI development. 

OpenAI CEO Sam Altman said of AI, “I think if this technology goes wrong, it can go quite wrong… We want to work with the government to prevent that from happening.” And Meta CEO Mark Zuckerberg said, “Congress should engage with AI to support innovation and safeguards.” 

Governments have responded with varying degrees of urgency. 

The European Parliament is leading the charge with their recent passage of the AI Act – the world’s first major set of regulatory ground rules to govern artificial intelligence. 

The UN, led by the United States, adopted its first resolution on AI shortly after, though President Biden is yet to make progress in Congress despite pushing for regulations. 

This environment of fear, uncertainty, and disjointed government response prompted us to investigate how US residents feel about governmental regulation of AI. 

Most Americans Want Strict AI Regulations

Our survey revealed a significant level of concern and a desire for stricter AI regulations. 

A striking 79.8% of respondents believed governments should implement strict AI regulations, even if it means slowing down technological innovation. This highlights the public’s prioritization of safety and ethical considerations over rapid advancement. 

When asked about the effectiveness of current regulatory frameworks in managing AI risks, 55.5% of respondents considered them effective, with only 14.55% regarding them as “very effective.” This suggests that while some progress has been made, there is still room for improvement in the eyes of the public.

Furthermore, 82.3% of respondents expressed support for either international standards (41.35%) or a mix of international standards and local laws (40.95%) in AI regulation, emphasizing the need for a coordinated global approach.

Privacy emerged as a major concern, with 82.45% of respondents expressing unease about the use of personal data in training AI systems. This figure rose to an alarming 91.96% among respondents in the IT and software industries, with 48.28% indicating the highest level of concern. This underscores the urgent need for robust data protection measures and transparency in AI development.

Lastly, an overwhelming 83.9% of respondents believed that AI companies should pay royalties for copyrighted content used in training AI models. This sentiment reflects the growing debate around intellectual property rights and reveals a desire for creators to receive fair compensation in the age of AI.

AI Regulation Analysis – Current State of the World

Our analysis reveals that nearly two-thirds of the world’s countries are actively working on regulating AI, with differing levels of progress.

To categorize countries’ progress on AI regulation, we developed a four-tier system:

• Comprehensive Regulation Enacted: Countries that have fully developed, enacted, and implemented comprehensive AI regulations. These regulations are broad, covering ethical concerns, privacy, transparency, security, and the societal impact of AI. This level signifies the highest degree of regulatory development.
• Active Regulatory Development: Countries with legislation in progress or those with frameworks or guidelines established. It includes nations actively working towards creating, revising, or finalizing AI-specific laws, as well as those that have set non-binding principles, frameworks, or guidelines to guide AI development and use.
• Initial Regulatory Efforts: Countries that have recognized the need for AI regulation and have begun preliminary efforts, such as research, stakeholder consultations, or the formation of advisory committees. This category highlights an early stage of engagement without substantial legal or formal guidelines.
• No AI Regulation: Countries that have not made discernible efforts towards AI regulation, either due to a lack of action, absence of information, or reliance solely on existing laws that do not specifically address AI.

Of 195 countries, 32 have already enacted comprehensive AI regulations, while 29 are actively working on implementing and finalizing laws. 53 countries are in the initial or drafting stages of their own AI-related regulations.

The European Union and China have taken the lead in AI regulation, albeit with different objectives. 

The EU’s AI Act prioritizes minimizing social harms through a comprehensive, risk-based approach, while China’s regulations focus on reasserting state control over information. 

Neither approach is likely to favor AI innovation, presenting a challenge for other nations to develop their own unique strategies that balance innovation, safety, and ethical considerations.

The United States and the United Kingdom are currently in the stage of active regulatory development. 

The UK government plans on targeted legislative interventions to address gaps in the current regulatory landscape, particularly concerning complex general-purpose AI systems. 

In the US, President Biden issued a landmark Executive Order to protect Americans from the potential risks of AI systems. 

However, a broader US response hasn’t come to fruition. While some US agencies have completed short-term actions and advanced longer-term directives, the progress of proposed bills in Congress remains uncertain.

Implications of Study

Our survey results show that Americans want strict, comprehensive, and globally coordinated AI regulations. Policymakers and industry leaders must address these concerns and create frameworks that ensure safety, privacy, and ethics while still allowing for innovation.

Our analysis of current AI regulations found that progress is being made, but much work remains. Only 32 out of 195 countries have comprehensive AI laws, meaning most nations are just starting to tackle this critical issue. The lack of a unified global approach makes it hard to ensure consistent standards and reduce risks.

The study also underscores the need for international collaboration in AI regulation. As AI crosses borders, countries must work together to establish global standards and best practices. This requires open communication, sharing knowledge, and a commitment to addressing AI challenges cooperatively.

AI Regulation Findings: Country-By-Country

Here are our complete findings from our analysis of 195 countries: