G7 performance on artificial intelligence, ICT and digitalisation
Share
G7 Summit

G7 performance on artificial intelligence, ICT and digitalisation

Nancy E Scott

Since June 2024, significant advances in artificial intelligence have transformed the digital landscape at lightning speed. We are witnessing momentous progress in the power and capability of generative AI, agentic AI and advances in areas such as quantum technologies, with growing energy demands resulting in an increased interest in nuclear energy, as well as clean and reliable energy. The G7 leaders’ focus on advancing the Hiroshima AI Process at their 2024 Apulia Summit resulted in continued commitments to advance responsible and secure AI development and deployment. Leading up to the 2025 Kananaskis Summit, former prime minister Justin Trudeau restated Canada’s commitment to advancing security, prosperity and partnership through AI adoption, energy and inclusion.

As the scale and strength of AI grow and expand across the globe, G7 leaders may struggle to advance a harmonised approach to supporting and developing responsible AI that reduces the digital divide and provides protection against harmful uses. The G7 has been a leader in supporting AI governance. With a growing risk of fragmented or inadequate regulation, the Kananaskis Summit is an important opportunity to continue to advance international AI governance.

Conclusions

From 1975 to 2024, G7 summits produced 23,729 words on digitalisation, information and communications technologies, and AI in their communiqués, averaging 475 words per summit (or 2% of the total). With some notable gaps (1979–1980, 1989–1993, 2003, 2005, 2008, 2010 and 2020), the topic was addressed at every summit since 1978. Between 2014 and 2024, the G7 dedicated an average of 1,348 words to digitalisation, ICT and AI per summit. At the 2024 summit, the leaders produced 1,825 words (9%), fewer than in 2023, when the leaders produced 3,927 words (for 13%) – the highest ever in absolute terms. The next highest was in 2000 with 3,020 words (22%) and 2021 with 2,170 (11%). By this measure, as host of the 2000 and 2023 summits Japan has led the G7 in its public deliberation of ICT, digitisation and AI.

Commitments

G7 summits have now produced 262 future-oriented, politically binding, collective commitments on ICT, digitalisation and AI. Collectively, these rank in 13th place, slightly behind macroeconomic policy (with 339 commitments) and before democracy (with 159 commitments). Most – 201 – have been made since 2016, averaging 22 per summit during these years. Notably, the first time AI commitments appeared was in 2018 at the Charlevoix Summit, where the G7 produced 23. The G7 produced 37 commitments on digitalisation, AI and ICT, in 2023, the highest ever, followed by 33 in 2021 and again in 2024.

Compliance

G7 members’ compliance with their commitments on ICT, digitalisation and AI averaged a very high 87%, based on the 19 priority commitments assessed by the G7 Research Group. This is well above the G7’s all-subject average of 77%. Compliance peaked at 100% with the commitments made in 2000, followed by 97% with those made in 2023. The lowest compliance was 72% for 2019 and 75% for 2018. By December 2024, compliance with the two assessed 2024 commitments, on AI for work and on closing digital divides, averaged 85%.

By member, compliance has been led by the United Kingdom at 95%, closely followed by the European Union at 94%, Canada at 92% and France at 90%. Then come the United States and Germany at 87%, Japan at 84% and Italy at 75%.

Recommendations

To improve the compliance with their commitments on ICT, digitalisation and AI, G7 leaders can take the following actions:

Make more commitments on AI, as these average compliance of 97%;

Make more commitments on countering foreign interference, especially from authoritarian regimes, using digital technologies, as such commitments also average compliance of 97%.

Make commitments that refer to a G7-centred institutional process or body, as such commitments average compliance of 96% (with that on the Dot Force at 100%, the Global Partnership on Artificial Intelligence at 94%, the Genoa Plan of Action at 88% and the Hiroshima AI Process at 100%). This suggests the value of having a related G7 ministerial meeting after the Kananaskis Summit to help implement the commitments made there and build on them in this specialised and fast-moving field.