The world needs an International Decade for Data–or risk splintering into AI ‘haves’ and ‘have-nots,’ UN researchers warn
The rapid rise in data-driven technologies is shaping how many of us live–from biometric data collected by our smartwatches, artificial intelligence (AI) tools and models changing how we work, to social media algorithms that seem to know more about our content preferences than we do. Greater amounts of data are affecting all aspects of our lives, and indeed, society at large.
This explosion in data risks creating new inequalities, equipping a new set of “haves” who benefit from the power of data while excluding, or even harming, a set of “have-nots”–and splitting the international community into “data-poor” and “data-rich” worlds.
We know that data, when harnessed correctly, can be a powerful tool for sustainable development. Intelligent and innovative use of data can support public health systems, improve our understanding of climate change and biodiversity loss, anticipate crises, and tackle deep-rooted structural injustices such as racism and economic inequality.
However, the vast quantity of data is fueling an unregulated Wild West. Instead of simply issuing more warnings, governments must instead work toward good governance of data on a global scale. Due to the rapid pace of technological innovation, policies intended to protect society will inevitably fall behind. We need to be more ambitious.
To begin with, governments must ensure that the benefits derived from data are equitably distributed by establishing global ground rules for data collection, sharing, taxation, and re-use. This includes dealing with synthetic data and cross-border data flows.
Although the digital economy has boomed in recent years, the digital and data divide has widened. While global internet bandwidth rose by roughly 35% in 2020 (a sign of greater connectivity and technological uptake worldwide), only 20% of people in the world’s least developed countries use the internet, and are typically doing so at significantly lower speeds and higher costs than in high-income countries. Poor and costly connectivity will handicap the efforts of these countries to participate in the dynamic global digital marketplace, currently estimated to be worth 15% of global GDP, and potentially increasing to 30% by 2030.
Global regulation is key. Current approaches are fragmented across regions and sectors and risk deepening prevailing economic divides, with the richest few benefiting most from data use while low-income economies are left behind, unable to take advantage of data to improve economic growth and livelihoods for their populations.
This deficit of global data governance also exacerbates the potential of real-world harm. Efforts to track terrorist and violent extremist experimentation with generative artificial intelligence have already uncovered over 5,000 instances of AI propaganda in support of violent and extremist ideologies.
Second, governments must ensure that more targeted data is being collected to support the development of traditionally underrepresented groups. Although it may seem like there is too much data, more–and better managed–data is needed to make technology responsive to and representative of underrepresented communities.
For instance, research into the potential biases of AI tools in healthcare–an issue of huge importance given AI’s ever-accelerating development–has been shown to suffer critically from a lack of diversity among its researchers. This lack of diversity in who is collecting data, and who for, risks engraining existing biases into new AI models over the long term, with disastrous outcomes for society at large.
To catalyze action, we need government leadership and space for constructive engagement with the private sector, multilateral institutions, and civil society. Governments can take the lead by championing a call for the launch of a UN International Decade for Data (IDD) starting in 2025.
International decades declared by the UN have proven relatively successful in coordinating action between states, the private sector, technical and scientific communities, academia, and civil society organizations to address issues of global concern.
For instance, the International Decade for Women (1976-1985) helped to galvanize action toward greater gender equality, launching the United Nations Development Fund for Women and led to the creation of a Convention on the Elimination of all Forms of Discrimination Against Women (CEDAW), now signed by 189 countries.
An International Decade for Data would establish a clear priority for the international community. It represents more than a political commitment to good governance of data–an International Decade can unlock funding to strengthen data capacities and nudge relevant stakeholders to collaborate, experiment, and regulate for a more equitable future.
The Global Digital Compact, to be agreed at the UN’s Summit of the Future in September 2024, presents an opportunity for governments to advance this global priority, expanding the benefits of data for all, while insulating against its potential risks.
AI is only the latest innovation in data-driven technologies–but it won’t be the last. The next 10 years will undoubtedly prove to be the data decade, with countless innovations that will fundamentally shape our lives. We need the political commitment that matches this opportunity.
Now is not the time for even more warnings. Governments and the UN must stay ahead of the curve of innovation. An International Decade for Data can help spur the international action needed to create response frameworks for data governance–allowing it to play the role of a global good, and not another future risk.
Tshilidzi Marwala is the rector of United Nations University (UNU). David Passarelli is the director of United Nations University Centre for Policy Research (UNU-CPR).
More must-read commentary published by Fortune:
The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.
Source link
