Artificial Intelligence and Foreign Policy

By Angela Cicily Joseph,

(Image source: Pixabay)

The Sputnik moment came when DeepMind’s artificial intelligence system AlphaGo defeated Go’s world champion, Lee Sedol by a marvellous 4 games to 1. Eighteen months later, AlphaZero was released by DeepMind. Unlike AlphaGo, human experts did not train AlphaZero. Instead, it taught itself and was able to not only defeat its predecessor but also other systems. Few months later, countries started investing and announcing national AI strategies.

While the race to develop artificial intelligence has gathered momentum with countries like China and the United States of America being leagues ahead, the other countries need to catch up for better economic and political prospects. With an automated world, the developing countries will have less demand for their typical supply of unskilled labour. With AI systems being trained to come up with policy recommendations, countries will have to prepare for a daunting political future.

The most popular example of this is the Cambridge Analytica scandal in which personal data of Facebook users was harvested without their consent to create psychographic profiles and influence political outcomes for US Senator Ted Cruz. Whistleblower Christopher Wylie has stated that during the Vote Leave campaign, a group of people were identified as persuadable and then were bombarded with false information to persuade them to vote for leaving the EU. When Facebook CEO Mark Zuckerberg was questioned by the members of the US Congress over this data scandal, it showed the lack of awareness and understanding among the Congresspeople. This is the crux of the matter. The lack of policy for AI stems from a place of obliviousness. Also, we do not know the extent and impact of AI interventions in our daily life.

Both China and the US have shown no indication to develop global AI policies, but they have taken up nationalist approaches in their strategies. With the development of AI-focused start-ups, the US has created an ecosystem for extensive AI research at government organisations and major companies like Facebook, IBM, Apple, Google, and Amazon. Unlike the consumer targeting at these companies that accesses data from some devices, AI will be able to draw data from multiple devices enabling predictive control of potential dissenters. China’s A Next Generation Artificial Intelligence Development Plan is one of the most comprehensive national AI strategies, announcing their ambition to become the primary hub for innovation by 2030. Expanding on the idea of credit checks, China’s Social Credit System enables individuals to rate each other’s behaviour and trustworthiness. A citizen’s rights can be curtailed if they do not pay their bills or even if they are found playing loud music. China and Zimbabwe have entered into an agreement to deploy facial recognition software to the capital city of Harare. Similar agreements have been signed with Angola and Ethiopia. China’s increasing investments in the continent of Africa should be a reason for concern.

While the largest players in the AI domain are China and the US, the global AI ecosystem has Israel, South Korea, Japan, UK, Russia and France with thriving AI sectors and massive investments.This growing technological inequality will have major geopolitical implications. This dominating narrative of an “arms race” introduces new risks; countries will be prompted to speed up their AI development. The danger is not about falling behind one’s competitors but engaging with unsafe AI programs. This rhetoric about an arms race must be ratcheted down and cooperation among countries must be the way forward.

(Angela Cicily Joseph is Research Associate at CPPR. Views expressed by the author are personal and need not reflect or represent the views of Centre for Public Policy Research)