Global Counsel releases latest original research on the regulation of generative AI
Global Counsel's latest research reveals limited anxiety amongst the public about immediate risks linked to proliferation of generative AI, in contrast to opinion formers, who voice deep concerns and desire to regulate; setting up challenging business environment for generative AI companies.
Global Counsel’s latest research report, ‘Regulating Generative AI’ has revealed a striking contrast in the perception of generative AI between the public and opinion formers: a low level of anxiety over the risks associated with generative AI amongst the public, contrasted by a deep resolve by opinion formers to encourage policymakers to act with urgency.
The findings serve as a reminder that policy elites may be primed to see their function as being to define the acceptable parameters for generative AI’s evolution. For businesses, the findings serve as a warning sign to not conflate a strong consumer brand with political brand. As social media companies before them have learned, a strong consumer brand will have limited impact with policymakers taking key decisions over regulation and safety.
Timed for release just ahead of the UK’s AI Safety Summit this November 1st and 2nd, the report’s findings are derived from extensive public polling of 3,000 individuals and interviews with 48 opinion formers drawn from government, regulatory bodies, industry and think tanks, across the US, UK and EU. The findings are intended to not only inform policymakers and industry leaders as they look to contribute to the latest global debate on AI governance and regulation, but also to anticipate policy change once the political landscape transforms following the UK, EU and US elections in 2024.
Conan D’Arcy, Senior Practice Director for Technology, Media and Telecoms at Global Counsel said:
“Our research confirms a number of key trends that will shape the regulation of generative AI and AI more broadly after the US, EU and UK elections next year. The first is that an international regulatory regime is unlikely or at best will be principles-based, meaning that national and regional rules, most obviously in the EU, will set global standards on AI regulation. The “Brussels Effect” is set to continue. The second is that, while Congress is unlikely to pass ambitious or comprehensive AI legislation, the US government is likely to pull on other levers, as we saw with the Executive Order this week, and US policy will continue to impact global AI supply chains through the restrictions it imposes on Chinese companies, regardless of who wins the White House next year. Finally, ahead of the UK’s AI Safety Summit, there is deep cynicism from government officials and regulators about the Summit’s focus on existential rather than near term risks from AI. They see it as a false dichotomy and believe governments globally should be mobilising to address both.”
The report finds that the debate over when and how to regulate is playing out against a backdrop of positive public attitudes to generative AI and little evidence of serious anxiety over its associated risks, at least when discussed in the abstract. Some findings from Global Counsel’s research includes:
- Unsurprisingly, younger adults and men had greater familiarity and favourable attitudes to generative AI across all three national surveys. Older people and women were less familiar and showed a less favourable attitude (women in the UK particularly so, with a -7% favourability rating).
- After acknowledging the wide range of risks after deliberation and debate in focus groups, the public’s view of the risks remained not well-established and rarely front-of-mind. Overall, US and German public were largely positive (+10 and +18% respectively), with UK public neutral (+0).
- Political association and favourability also correlated. In Germany and the US, there appears to be more correlation between support for traditional ‘mainstream’ parties and greater favourability toward AI – 2020 Biden voters have a far more favourable view of generative AI than 2020 Trump voters (+24% to +1% respectively).
- Beneath this basic consensus lay important nuances and differences, though robust support for regulation by the public. Those who were more familiar with and favourable towards generative AI in general (particularly younger adults and men) also tended to favour less significant regulatory restrictions. German citizens were less likely than their British and American counterparts to favour heavy restrictions.
- There was also some variation in attitudes between use cases. For example, whereas 29% of US adults were in favour of banning entirely the use of generative AI to run social media accounts, only 10% supported a ban for use in fraud detection and prevention.
- For businesses, there is an onus to demonstrate the relevance of generative AI use cases to salient policy challenges faced by governments. This is clear in the research in respect to healthcare and financial services applications, two already highly regulated sectors where existing frameworks may suffice to tackle near term risks.
Several events are planned to socialize the findings of the report, including a private roundtable launch event in London in the lead up to the UK’s AI Safety Summit, and a launch of the report in partnership with the U.S. Chamber of Commerce in Washington DC. The findings will also be covered in an online event this November 3rd, immediately following the London summit.
About Global Counsel’s Technology, Media and Telecoms Practice
Global Counsel’s Technology, Media and Telecoms (TMT) team provides clients across multiple markets with policy analysis, political insight, and strategic advice on a range of cross-cutting issues such as digital competition, data protection, digital infrastructure and online safety, as well as on emerging technologies like generative AI and virtual worlds. They provide expert advice to clients to navigate the complex regulatory and political landscape, anticipate market trends and develop successful commercial and policy strategies. Each year the team publishes original research unpacking the regulatory issues and trends which will impact global technology policy and wider implications for the political economy.