New EU Act Aims To Keep AI Safe
The world’s first major Act to regulate artificial intelligence passed into law across the European Union in August. It aims to ensure the technology is safe and respects the bloc’s fundamental rights and values. This legislation is set to have a significant impact on businesses operating within the EU.
Concerns About AI Development
The meteoric rise of AI has sparked countless dystopian scenarios, raising deep concerns about the race to develop increasingly advanced systems. A notable open letter signed by influential tech leaders, including Elon Musk and Steve Wozniak, called for a pause in the development of advanced AI models. The signees warned that “AI systems with human-competitive intelligence can pose profound risks to society and humanity.”
Business Unease
Businesses are expressing unease regarding the potential dangers of AI, inhibiting its uptake. A poll of Irish business leaders found that 91% believed generative AI would increase security risks in the upcoming year, while 53% feared catastrophic cyberattacks. Furthermore, almost three-quarters thought that such tools would “fail to enhance” their company’s ability to build trust with shareholders.
Other concerns highlighted include legal liability, reputational risks, and the spread of misinformation. Most critically, less than three out of ten respondents were confident in their organization’s current measures to control generative AI for safe outcomes.
New Regulatory Framework
The EU Act introduces a risk-based approach to AI regulation. The classifications are as follows:
- Unacceptable risk: AI systems that threaten fundamental rights are banned. This includes models used for manipulating human behavior or social scoring.
- High risk: Systems in critical sectors such as healthcare or public administration must comply with strict data quality and transparency standards.
- Limited risk: Measures to promote trust and ensure users are aware they are interacting with AI systems.
- Minimal risk: Applications posing little or no threat, such as spam filters, can be deployed with fewer regulatory requirements.
The legislation will be enforceable from August 2, 2026, with some rules applying sooner. For instance, the ban on systems presenting an unacceptable risk will take effect on August 1.
Support for Innovation
To support innovation, the EU has allocated €4 billion in funding for startups and small to medium-sized tech firms to develop compliant AI models. Digital innovation hubs will provide personalized training programs and mentorship to businesses of all sizes transitioning to AI technologies.
Implementation and Monitoring
AI system providers will be required to have ongoing monitoring regimes in place, with serious incidents needing to be reported. EU member states must appoint watchdogs to enforce compliance with the legislation and conduct market surveillance by August 2 next year.
The Future of AI Regulation
The EU’s approach emphasizes that AI has the potential to change the way we work and live, promising enormous benefits for citizens and the economy. However, the successful implementation of this legislation remains a challenge, with concerns that it could create regulatory barriers for European companies.
As the AI landscape continues to evolve, the focus must shift from legislation to practical application, ensuring that the technology serves humanity positively while adhering to established rights and values.
