In today’s world, AI can be a recipe for success – but don’t forget that data is its key ingredient

The world today is characterised by countervailing forces: a borderless digital economy is advancing like never before, while the international system is descending into a deepening recession. Geopolitical divisions are bringing fragmentation across the traditional economy, from global supply and value chains to international flows of capital, and increasingly in the digital sphere.

While global leaders have taken proactive steps to provoke agreement on the most pressing issues of the moment – the UK’s initiative in convening the AI Safety Summit at Bletchley Park in 2023 is a pertinent example – cooperation on digital and technology policy remains on precarious footing. With elections taking place in more than 60 countries in 2024, bringing potential new governments, with new mandates and new priorities, the tightrope could shake again.

Understandably, given the profound – and sometimes existential – questions attached to the issue, artificial intelligence (AI) has been prioritised by policymakers around the world. The collective need to manage the biggest threats and opportunities that this technology presents is a source of common ground across borders. But with a predominant focus on AI’s possible effects, there’s a risk that we lose sight of collaborating to ensure the integrity of the key ingredient that goes into it: data.

Since long before the explosion of AI, data has been the lifeblood of the global economy and financial system. Flowing from sources to users around the world, data underpins how we all work and live – whether it’s about accessing the cloud, making informed strategic decisions or facilitating remote working, its enabling role is ubiquitous in organisations of all sizes, sectors and geographies. With services representing an increasing proportion of the global trade mix alongside goods, cross-border data flows are only becoming more important.

LSEG alone delivers around 300 billion data messages to customers across 190 markets every day. With data becoming ever more central to the global economy, we have seen customer demand for our data rise by around 40% per year since 2019.

To unlock the value of the data to our global economy and society, it needs to be able to flow across borders freely, securely, and with trust. For example, globally accessible know-your-customer data helps financial institutions to prevent their business from being used to launder the proceeds of financial crime, and cross-border data flows allow firms to monitor and respond to cybersecurity and other operational risks in real time.

When Russia invaded Ukraine, LSEG saw a 600% spike in customers’ use of our risk intelligence database to implement sanctions quickly and effectively. With 40,000 requests from customers per day, the ability for data to cross borders makes a huge difference in our ability to meet their needs.

The need for trusted data flows is no different when it comes to capitalising on the opportunities and minimising the risks associated with AI. An AI system’s output is only as good as its data inputs – its potential depends on access to large datasets sourced from jurisdictions around the world, and robust processes to ensure the quality, integrity and lineage of data.

This is particularly apparent in the financial services sector, which is rapidly adopting AI systems across a range of use cases. Our industry relies on pinpoint accuracy: if LSEG were to tell our customers that our data and analytics were 95% accurate, it would simply not be good enough.

In this respect, there’s certainly a need to coordinate on policy that promotes the responsible use of AI, and we’ve seen welcome progress through fora such as the G7 and the OECD. LSEG has drawn on this work to inform our own approach in developing Responsible AI Principles.

But continuing to advance cooperation on data governance and strong digital trade rules is also an important part of the recipe. The G7 and OECD have been central on this front too, driving international cooperation to promote cross-border data flows while ensuring trust in privacy, security, and intellectual property rights. The Japanese Government was pivotal in gaining political support for “Data Free Flow with Trust” (DFFT) during its 2023 G7 Presidency, and now the OECD is leading multi-stakeholder work to operationalise this agenda at a technical level.

DFFT has many applications, and enabling the development of robust, reliable and responsible AI systems that bring benefits to the economy and society is one of them. Looking ahead, policymakers should consider these issues holistically, connecting the dots between parallel workstreams that are intrinsically linked, and continuing to go further and faster at pace with the rapid technological transformation that the world is undergoing.

You can’t make a good meal without the right ingredients – and for AI to meet its potential in addressing the world’s biggest challenges, we need to be able to rely on the data that’s going into it.