This alliance aims to accelerate the adoption of inclusive, trusted and transparent AI worldwide

This article is brought to you thanks to the collaboration of The European Sting with the World Economic Forum.

Author: Kay Firth-Butterfield, Head of Artificial Intelligence and Machine Learning; Member of the Executive Committee, World Economic Forum & Mark Caine Project Lead, Artificial Intelligence and Machine LearningProject Lead, Artificial Intelligence and Machine Learning, World Economic Forum

• AI could promote social good – if we address the risks.
• The World Economic Forum’s The Global AI Action Alliance brings together a multistakeholder group including more than 100 companies to accelerate the adoption of inclusive, trusted and transparent AI.

Artificial intelligence (AI) could contribute more than $15 trillion to the world economy by 2030, adding 14% to global GDP, according to PwC. This would make AI the biggest economic opportunity of the next decade, representing more value than today’s insurance, oil and gas, commercial real estate, and automotive industries combined.

Along the way, AI could bring huge benefits to society. A 2018 study by Google identified 2,602 AI use cases that promote social good, and people are increasingly applying AI to address critical societal challenges including improving agricultural yields, reskilling workers and combatting COVID-19.

But as the economic and social potential of AI has become clear, so, too, have the risks posed by unsafe or unethical AI systems. Recent controversies on facial recognition, automated decision-making and COVID-19-related tracking have shown that realizing AI’s full potential requires strong buy-in from citizens and governments, based on their trust that AI is being built and used ethically.

Only robust collaboration can ensure that we maximize the benefits of AI and distribute them equitably across society. —Kay Firth-Butterfield and Mark Caine

Many companies and governments have recognized this, and progress has been made in defining ethical principles for the development and use of AI – to date more than 175 separate sets of principles have been proposed.

While the explosive growth in AI ethics guidelines is welcome, it has created an implementation gap – it is easier to define the ethical standards a system should meet than to design and deploy a system to meet them. It has also created learning gaps, with promising activities happening in silos and few mechanisms in place to drive learning, action, and impact at scale.

Unfortunately, it’s not just AI ethics efforts that are happening in silos – so too are successful implementations of AI that capture its economic and social benefits and distribute them equitably across society. While AI holds the promise to improve billions of lives, the reality to date is that the AI ecosystem has underserved the majority of global citizens.

This reflects in part its lack of inclusivity: according to New York University’s AI Now Institute, just 18% of authors in leading AI conferences are women, roughly 80% of AI professors are men, and non-white engineers comprise less than 5% of most major technology companies’ workforces. And if current AI adoption trends continue, North America, Europe, and China will capture roughly 80% of the economic benefits brought by AI, leaving just 20% for the remaining two-thirds of the global population.

This is a recipe for exacerbating global inequality and fuelling socio-economic instability, not to mention a huge missed opportunity to uplift billions of lives and improve the state of the world. It’s little wonder that many people are scared of the future that AI might bring: according to recent studies by Oxford University and the European Commission, 84% of Americans and 88% of Europeans believe that AI should be “carefully managed” on the basis of its potential to result in harmful societal outcomes.

With AI’s impact on industry and society accelerating every day, and with so much at stake in how its roll-out is managed, there is an urgent need for greater collaborative effort to ensure that it is used ethically and in the global public interest.

Coming together for inclusive, transparent AI

That is why the World Economic Forum is launching The Global AI Action Alliance, a new multi-stakeholder collaboration platform and project incubator designed to accelerate the adoption of inclusive, trusted and transparent AI globally and across industry sectors.

The alliance brings together more than 100 leading companies, governments, international organizations, non-profits, and academics united in their commitment to maximizing the societal benefits of AI. Alliance members work together to identify and implement the most promising tools for ensuring that AI systems are ethical and serve all of society, including groups historically underserved by AI.

The alliance provides a platform for members to engage in real-time learning, pilot new approaches to ethical AI, scale adoption of best practices, and undertake collective action to ensure that AI’s benefits are shared by all. Supported by a grant from the Patrick J. McGovern Foundation, the alliance is overseen by a Steering Committee of top global leaders from industry, government, academia and civil society and co-chaired by Arvind Krishna, Chairman and CEO of IBM, and Vilas Dhar, President of the Patrick J. McGovern Foundation.

No single organisation can address the full range of challenges presented by AI, nor can any one actor deliver the immense benefits that AI can offer to society. With so many challenges to overcome and so many opportunities to unlock, only robust collaboration can ensure that we maximize the benefits of AI and distribute them equitably across society.

The costs of inaction are high, but the benefits of action are even higher. That is why the time is now for global action to create a positive, human-centred future for AI.