The Shift to a Softer Approach on AI Regulation

Exploring the current state of innovation and safety paradigms in AI Development

The political and economic winds of change blowing from the White House have seemingly upended the direction of travel for AI policy over the last few months.  A starting point here is perhaps the Executive Order issued merely three days into the new US President assuming office titled “Removing Barriers to American Leadership in Artificial Intelligence”[1]. This expeditious executive action from the Trump administration repealed a Biden era Executive Order which mandated the safe, secure and trustworthy development and use of Artificial intelligence.[2] Instead, its stated goals include ridding the AI development lifecycle of “ideological bias or engineered social agendas” and solidifying the USA’s position as a global leader in AI.

Another full-throated warning arrived with Vice President JD Vance’s speech at the Munich Security Conference this year warning the EU to adopt an approach to foster the creation of AI technology rather than “strangle it”.[3] More recent news from the US on AI policy has ranged from calls to  “delete” intellectual property laws for impeding AI development[4], diplomatic pressure to ditch the General Purpose AI Code of Practice[5], and the firing of the Head of the US Copyright Office soon after releasing a report which amongst other research, indicated that some AI development might violate fair use provisions in copyright laws. [6]

Given the indirect positive economic effects of AI and the fact that AI is increasingly seen as a national security asset with crucial applications in military and security, the EU has taken some key steps to rethink its strategy. The Commission’s own data indicates that venture capitalists invested about $8 billion in artificial intelligence in the EU, compared to $68 billion in the U.S. and $15 billion in China.[7] Unsurprisingly then, the recent Paris AI Action Summit saw Henna Virkunen, the European Commission’s lead voice on technology stating that the EU AI Act should not be creating “any extra burden” on tech companies and highlighted the need for a revision of the AI Act reporting obligations to enable a simplified implementation and support structure.[8]

Additionally, the European Commission has also launched a consultation regarding the development of the Apply AI Strategy to serve as a blueprint for the adoption of AI in EU strategic sectors to make Europe an ‘AI continent’. The consultation specifically calls out that “Most developments in AI are currently taking place in non-EU countries” and deems the private investment and AI adoption by mid-caps and SMEs in Europe as insufficient.[9]

Europe’s rights oriented and humanistic approach traditionally served as the counterweight to US’ pro-innovation approach. However, this balancing act is now at risk of being obliterated in view of the importance of AI as a force multiplier, geopolitically and economically. This has resulted in certain AI models and capabilities being launched in most major markets leaving the EU citizens deprived.[10]

As this tug of war plays out, the risks of poorly regulated AI are only increasing at breakneck pace as the technology progresses from hype to business reality and businesses seek to avoid commercial and innovation irrelevance.[11] However, recent research suggests that if social damages are large and proportional to the new technology’s productivity, a higher growth rate paradoxically leads to slower optimal adoption.[12]  In this backdrop, it becomes important, for us a society, to consider what the ideal speed for innovation should be with technologies that might also have deleterious effects.

*The views expressed in this article are solely personal and independent. They do not represent the views of any employment or organizational affiliates that the author might have.

Any article or other information or content expressed or made available in this Section is that of the respective author(s) and not of the OBA.