The riskiness of AI development results from the fact that modern AI tools are pushing ethical boundaries under the existing legal frameworks that weren’t made to fit them. However, the way regulators choose to proceed in light of this varies greatly between different countries and regions.
The recent AI Action Summit in Paris highlighted these regulatory differences. Notably, its final statement focused on matters of inclusiveness and openness in AI development while only broadly mentioning safety and trustworthiness and without emphasizing specific AI-related risks, such as security threats or existential dangers. Drafted by 60 nations, the statement was not signed by either the US or the UK, which shows how little consensus there is in this space.
Head of Product Legal Counsel at Oxylabs.
How different regulators tackle AI risks
Countries and regions differ from each other in how they regulate AI development and deployment. Nonetheless, most fit somewhere between the two poles constituted by the United States at one point of the extreme and the European Union at the other.
The US way: innovate first, regulate later
The United States has no federal-level acts regulating AI in particular; instead it relies on voluntary guidelines and market-based solutions. Key pieces of legislation include the National AI Initiative Act, which aims to coordinate federal AI research, the Federal Aviation Administration Reauthorization Act and the National Institute of Standards and Technology’s (NIST) voluntary risk management framework.
In October 2023, President Biden issued an Executive Order on Safe, Secure and Trustworthy Artificial Intelligence, establishing standards for critical infrastructure, enhancing AI-driven cybersecurity and regulating federally funded AI projects. However, the US regulatory landscape remains fluid and subject to political shifts. In January 2025, President Trump revoked Biden’s executive order, signaling a potential pivot towards promoting innovation and away from regulation.
Criticisms of the US’s approach include its “fragmented nature,” leading to a complex web of rules, “lack of enforceable standards,” and “gaps in privacy protection.” However, the country’s laissez-faire approach to regulating AI may very well change in the future. For instance, in 2024 alone, state legislators introduced almost 700 pieces of AI legislation and there have been multiple hearings on AI in governance, AI and intellectual property, etc. This shows that the US government doesn’t shy away from regulation but is looking for ways of implementing it without compromising too much on innovation.
The EU way: damage-prevention approach
The European Union has taken a very different approach. In August 2024, the European Parliament and Council introduced the Artificial Intelligence Act (AI Act), widely regarded as the most comprehensive piece of AI regulation to date. Employing a risk-based approach, the act imposes the most stringent rules on high-sensitivity AI systems, e.g., those used in healthcare and critical infrastructure, while low-risk applications face only minimal oversight. Certain applications, such as government-run social scoring systems, are explicitly prohibited.
Similar to the GDPR, the act mandates compliance not only within the EU’s borders but also from any provider, distributor, or user of AI systems operating in the EU or offering AI solutions to its market (even if the system has been developed outside it). This may pose some challenges for US and other non-EU providers of integrated products. Criticisms of the bloc’s approach include its alleged failure to set a gold standard for human rights, excessive complexity and lack of clarity and highly exacting technical requirements at a time when the EU is seeking to bolster its competitiveness.
Regulatory middle ground
The United Kingdom has adopted a “lightweight” framework somewhere between the EU and the US. The framework is based on certain principles and core values like safety, fairness and transparency. Existing regulators, such as the Information Commissioner’s Office, are empowered to implement these principles within their respective domains.
In November 2023, the UK founded the AI Safety Institute (AISI), evolving from the Frontier AI Taskforce. AISI is tasked with evaluating the safety of advanced AI models, collaborating with major AI developers to conduct safety tests and promoting international standards. The UK government has also published an AI Opportunities Action Plan, outlining measures to invest in AI foundations, implement cross-economy adoption of AI and foster “homegrown” AI systems. Criticisms of the UK’s approach to AI regulation include limited enforcement capabilities (“all eyes, no hands”), a lack of coordination between sectoral legislation and a lack of a central regulatory authority.
Other major countries have also found their own place somewhere on the US-EU spectrum. Canada has introduced a risk-based approach with the proposed Artificial Intelligence and Data Act (AIDA), designed to balance innovation with safety and ethical considerations. Japan has emphasized a “human-centric” approach to AI, publishing guidelines to promote trustworthy development.
In China, AI regulation is tightly controlled by the state, with recent laws requiring generative AI models to align with socialist values and undergo security assessments. Australia, meanwhile, has released an AI ethics framework and is exploring updates to its privacy laws to address emerging challenges.
Establishing international cooperation
As AI technology continues to evolve, the differences between regulatory approaches will become even more apparent. Whatever the individual approach certain countries take regarding data privacy, copyright protection and other aspects, a more coherent global consensus on key AI-related risks is badly needed. International cooperation is crucial in establishing baseline standards that both address key risks and foster innovation.
Currently, global organizations like the Organisation for Economic Cooperation and Development (OECD), the United Nations and several others are working to establish international standards and ethical guidelines for AI. The path forward requires everyone in the industry to find some common ground. And if we consider that innovation is moving at light speed, the time to discuss and agree is now.
We’ve listed the best laptop for computer science students.
This article was produced as part of TechRadarPro’s Expert Insights channel where we feature the best and brightest minds in the technology industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro
