Building an AI economy that works for all

Artificial intelligence is reshaping economies at remarkable speed, with some estimates placing the value of the current global AI market at $750 billion, exceeding the gross domestic product of most countries.

While this is an impressive figure, history shows that market share does not always lead to shared benefits. 

And without thoughtful guidance and dialogue, the promise of AI risks bypassing many around the world who would otherwise benefit most.

That is why AI governance, built on collaboration, shared principles and responsible stewardship, was a central theme of the International Telecommunication Union’s 2025 AI for Good Global Summit, which brought together 169 countries and over 10,000 participants.

Despite different national and regional approaches, we heard clear opportunities to steer AI development in a positive direction by focusing on three key priorities: inclusion, capacity building and technical standards.

1. Inclusion must be at the heart of
AI governance efforts. 

All countries and communities must have a voice – and the means – to participate in shaping our shared digital future.

That begins with universal connectivity – and yet 2.6 billion people remain offline. 

Across the African continent, where the G20 summit and presidency are being hosted for the first time, just 38% of the population has internet access, and mobile broadband can cost up to 14 times more than in Europe. 

Infrastructure access also remains limited, with just 32 countries hosting AI-ready data centres, while more than 150 – mainly in the Global South – have none.

AI can only expand economic opportunities for all if every country has a seat at the governance table – not just those with the most computing power or resources.

2. Infrastructure must be matched with capacity to use AI effectively. 

This is especially true in the public sector, where decisions can shape services for millions.

For AI to truly deliver for everyone, people from all socio-economic backgrounds must be able to participate in the full AI and data value chain. 

Because countries vary widely in readiness, education and skills programmes are essential to support better jobs and enable communities to shape AI to local needs.

3. Technical standards are key to transforming governance commitments into operational safeguards.

Standards help address risks such as environmental impact, bias and disinformation, while also creating economies of scale and levelling the playing field, so that companies of all sizes can share in a burgeoning AI economy.

Standards can help make AI systems interoperable, adapt them to local contexts and embed sustainability across the AI value chain – from sourcing raw materials cleanly to running data centres on green energy. 

What ITU is doing

As the United Nations agency for digital technologies, ITU works across all three areas to ensure digital technologies – including AI – reach and uplift every community.

As a proud knowledge partner of the G20 Digital Economy Working Group and contributor to the AI Task Force, ITU advocates for the fair, inclusive and sustainable development of artificial intelligence, including through our AI Governance Dialogue: a neutral, global platform where all countries – regardless of GDP or geography – can discuss policy approaches on an equal footing.

Through our AI Skills Coalition of over 40 partners, ITU also offers training opportunities to expand global talent pipelines, especially in emerging economies.

And as a standards development organisation, ITU collaborates with ISO and other partners, including the private sector, to build the technical foundations for sustainable AI. 

The recently launched AI Standards Exchange compiles 700 standards and publications into a single database designed to help companies, policymakers and regulators ensure that AI is trustworthy, interoperable and resilient.

What G20 members can do

Representing two-thirds of the world’s people and 85% of global GDP, G20 members have the capacity – and responsibility – to lead. 

As leaders meet in Johannesburg, they can:

• Close the digital divide and prevent an AI divide by investing in affordable, meaningful connectivity.
• Participate in inclusive global AI governance dialogues.
• Invest in capacity building and AI skills development initiatives.
• Contribute expertise to international AI standards processes.
• Use milestones such as the upcoming 20-year review of the World Summit on the Information Society to renew global commitments to inclusive digital
  transformation.

Sustainable economic growth in the AI era depends on technologies serving everyone, not just the best resourced. 

Inclusion, capacity and standards are three foundational building blocks that can bring us closer to that vision.

Our digital future is ultimately shared – and only by working together can we ensure AI delivers for all of humanity.