Ukraine and China will dominate G-7 summit, but a new threat lurks: AI

HIROSHIMA, Japan – President Joe Biden began his foreshortened Asia trip on Thursday in Hiroshima, a city that devotes itself to reminding the world of what happens when a brutal war escalates into a nuclear one.

There he prepared for discussions with his closest allies on two crucial issues: how to better arm Ukraine as it enters its counter-offensive against the Russian invaders, and how to slow, or halt, the downward spiral in relations with China.

Both are now familiar topics to the leaders of the Group of 7 nations, who have grown far tighter, and have remained surprisingly unified, since Russia began its assault on Ukraine 15 months ago.

But at some point over three days of discussions, the G-7 leaders are also expected to venture into new territory: the first conversations among the world’s largest democratic economies about a common approach to regulating the use of generative artificial intelligence programs such as GPT-4.

Artificial intelligence was not on the early agenda as Prime Minister Fumio Kishida invited the other six leaders – joined by Prime Minister Narendra Modi of India and, via video or in person, President Volodymyr Zelensky of Ukraine – to the Japanese prefecture where he got his political start.

But as the new artificial intelligence language model from OpenAI made nations around the world focus for the first time on the possibilities for disinformation, chaos and the physical destruction of critical infrastructure, Mr Biden’s national security adviser, Jake Sullivan, began calling counterparts to seek a common discussion.

It is far from clear that this group of leaders – the G-7 also includes Germany, Britain, France, Canada and Italy – can sustain a conversation on a technology that appeared to burst on the scene so quickly, even if it was years in the making.

Past efforts to get the group to take up far more straightforward cybersecurity issues usually descended into platitudes about “public-private partnerships,” and there has never been serious discussion of rules to guide the use of offensive cyberweapons.

US officials say that in the case of chatbots, even a vague foundational discussion may help in establishing some shared principles: that the corporations that bring products using the large-language models will be primarily responsible for their safety, and that there must be transparency rules that make it clear what kind of data each system was trained on.

That will enable lower-level aides to discuss details of what those first regulations would look like, the officials said.

But as the G-7 leaders convene starting Friday, it will be Ukraine that will dominate the conversation, at a critical moment for Mr Zelensky, for Ukraine and for the core Western democracies now seized with an urgent mission of bringing about what Biden calls the “strategic defeat of Russia in Ukraine”.

Some of the core members are seeking to arm Mr Zelensky in ways that may outpace Mr Biden’s willingness.

When he was in Britain, Prime Minister Rishi Sunak embraced Mr Zelensky in a bear hug and told reporters, “They need the sustained support of the international community to defend against the barrage of unrelenting and indiscriminate attacks that have been their daily reality for over a year. We must not let them down.”