YouGov Plc has recently conducted two polls commissioned by Control AI. One poll was on the public in France, Italy, and Spain, and the other was on the UK public. The members of the public selected for each country evenly represented all age ranges, political party allegiances, and sexes.
The polls asked a range of questions to the public concerning AI regulation and perceptions of the dangers posed by different types of AI. The public’s response to poll questions revealed strong majority support for regulation of AI creation and AI usage.
The polls found near universal support for either banning deepfakes outright, or at least severely regulating them. A summary of these results:
In the United Kingdom, 86% of the public support a ban on non-consensual deepfakes, with only 5% not in support.
In France, 77% of the public support a ban on non-consensual deepfakes, with only 10% not in support.
In Italy, 74% of the public support a ban on non-consensual deepfakes, with only 12% not in support.
In Spain, 82% of the public support a ban on non-consensual deepfakes, with only 11% not in support.
In all countries support was non-partisan, spread among all age demographics and both sexes. With such extensive support, it is surprising to see that creation of non-consensual deepfakes remains legal across all of these countries.
Furthermore, the public also expressed a distrust in AI companies to create safe technology for the public benefit. In Italy, only 28% trusts AI companies. In France, trust is even lower, with only 19% of the public trusting AI companies. In the United Kingdom, trust in AI companies is lowest. Only 12% of the public trusts AI companies.
It is no surprise that there is a lack of trust in AI companies. The technology that many of these companies create is currently used freely by malicious actors to create non-consensual deepfakes. Much worse than this, high executives at AI companies themselves admit that AI has the potential to collapse civilization, or even to kill us all. Yet they are still recklessly scaling, creating more powerful models in a race towards godlike AI.
The lack of trust in AI companies generally translates into a desire for AI companies to be externally regulated. 69% of the UK public believe that state-of-the-art AI should be regulated by the government. Regulating at this level would not only greatly reduce the existential risk posed by AI. It would also make it much harder for malicious actors to create non-consensual deepfakes.
To truly tackle the dangers posed by deepfake technology, we must regulate the technology itself, rather than just those who use it.
All figures, unless otherwise stated, are from YouGov Plc. Total sample size was 1,095 adults from France, 1136 from Germany, 1110 from Spain, 1052 from Italy, and 2086 in the UK. Fieldwork was undertaken between 21st - 28th November and 18th-19th October 2023. The surveys was carried out online. Each survey sample has been weighted and is representative of all adults in its respective country (aged 18+).