AI will test faith in democracy, Tokyo warns

Democratic governments in the US, UK and Japan will soon face a series of pivotal showdowns pitting the trust of the general public against the potentially “very disruptive” powers of generative artificial intelligence, Japan’s digital minister has warned.

In an interview with the Financial Times ahead of the G7 leaders’ meeting in Hiroshima, Taro Kono cautions that forthcoming general elections in Britain and Japan, as well as the US presidential race next year, could become ripe targets for “malicious elements” empowered with AI.

US voters, he says, have already experienced disinformation campaigns during presidential elections, but those were primarily carried out by human beings. “If the same thing can be done by AI, the sheer volume would be humongous,” says Kono.

“All governments need to consider how we can keep the trust of the people towards democracy . . . all democratic governments now feel an urgency in dealing with AI, so that’s why, at the G7, it is on everyone’s mind,” he explains, noting that the subject had been discussed extensively during preparations for the various leadership summits taking place this year.

Japan’s leadership of the G7, says Kono, has coincided not only with the release of the AI chatbot ChatGPT last November, but with a tectonic shift in the conversations surrounding technology. He suggests that, at the January meeting of the World Economic Forum in Davos this year, it was striking how the term “Web 3.0” — once a favourite catch-all for discussions around disruptive technology — completely vanished.

Instead, he says, everyone was talking about AI because suddenly ChatGPT had created something tangible where, previously, the conversation had been more theoretical. People could use ChatGPT, he says, and directly feel the power of what AI can do. “You can see what to expect in the future,” he says.

But, in parallel with the grand tech-related questions raised by advances in AI, the G7 will be having a number of more immediately practical discussions, too. In particular, Japan has pushed hard to include a discussion on cross-border data flow — a subject on which the US and Europe are sharply divided in their approach to regulation, and where Kono believes Japan has been able to provide a bridging function.

Democratic governments now feel an urgency in dealing with AI . . . at the G7, it is on everyone’s mind

While international bodies such as the World Trade Organization and Financial Stability Board exist to monitor their respective policy areas, no such body exists to oversee the global flow of data, despite its now crucial importance to the modern economy.

At Davos, in 2019, Japan’s late prime minister Shinzo Abe proposed the idea of an international order described as Data Free Flow with Trust (DFFT), which aimed to reconcile economic and privacy considerations. At this year’s gathering of digital ministers in Gunma, the G7 endorsed the establishment of a non-binding institutional framework — potentially housed at the OECD — that industry groups say would serve as a first step towards global data governance.

But the problem that the scheme quickly encountered was the highly fragmented patchwork of regulatory regimes. Europe’s legislation, which came into effect in 2018, is at the toughest end of the regulatory scale. By contrast, the US, says Kono, “is the wild, wild west, where anything goes”. Japan, he says, is probably closer to the US position, and would instinctively tend to wait for emerging technology to develop by some distance before working out whether it needed regulation.

“So it is quite difficult to get convergence,” says Kono. “I mean, we’re not going to have one set of rules concerning data transfer globally in a very short time.” Still, while it is important to see how far apart everyone is, he also believes that it is worth stressing the need for increasing “interoperability” on the issue.

The final communique of the G7 is expected to include agreement on the principle that, while the cross-border flow of data and information is positive for productivity and innovation, it raises significant challenges related to privacy, protection of intellectual property rights and security.

According to people familiar with the document, the draft of the communique states the G7 nations’ intention to operationalise the concept of DFFT, with a view to gaining more widespread support for the principle and greater convergence between existing regulatory approaches.

Japan’s attempt to take a leadership role on this issue of data flow coincides with a period in which the country is struggling to secure a reputation on all things digital. Kono — an experienced former foreign minister and leadership challenger — became Japan’s minister for digital affairs in August 2022. He inherited, with enthusiasm, a project to drag Japan’s paper-based bureaucracy into the digital age but has continued to encounter powerful resistance.

The mismatch between Japan’s economic strength and its digital prowess can be striking. As foreign minister, Kono was used to making offers of assistance to his counterparts from around the world.

“I would tell them that, if there was anything Japan can do, we are happy to give you assistance. But ever since I became digital minister, it’s my counterpart offering anything they can do to help Japan’s digitisation. Everyone knows we are lagging behind,” he says.