By Javier Surasky-
Countries
are giving AI a marginal place when considering their sustainable development
paths, which is a severe mistake. World leaders talk much about AI, its
risks, and its potential to bring significant benefits. But do they truly
understand them?
When
analyzing the Voluntary National Reviews (VNRs) presented by countries to the
2024 High-Level Political Forum, we see that countries need help integrating AI
when charting their trajectories towards sustainable development.
It seems
strange to make this statement just when AI has firmly established itself on
international agendas, even at the United Nations: it has been a constant
reference in countries' speeches during the 2023 SDG Summit and in the speeches
given at the last General Debate of the United Nations General Assembly, which
has adopted its first two resolutions on AI in recent months.
That was
the reason that led us to analyze how the VNRs presented in 2024 included AI,
and the result leaves us concerned. We worked on a universe of 28 reports with
official versions in English or Spanish, representing 80% of the shared VNRs
(36 reports in total). The list of studied VNRs is at the end of this post.
Limiting
ourselves then to our study universe, we found that only eight reports include
some consideration of AI, which is equivalent to just under a third (32%). Of
these, only two give a cross-cutting treatment of AI throughout the VNR:
Austria and Mauritius. The rest only consider AI sporadically and
discontinuously, which also occurs in Spain's report despite being the only one
that frames AI in policy coherence for sustainable development (p.156-157).
The
development of AI competencies is the most repeated topic among the mentions of
artificial intelligence. Considering this fact and remembering that AI is not an end but a means, I have tried to understand why countries want to invest in
creating AI competencies.
Among the
most mentioned topics are:
- The alignment between people's skills and market needs.
- Agricultural development and food production.
- Improvement in health systems and disease treatments.
Less
expanded than the previous issues, it is still possible to find references to
improving transparency and accountability of public institutions, gender equity
in the development and use of AI, and promoting international cooperation to
reduce IA gaps. I have not found any mention that links South-South Cooperation
or Triangular Cooperation with AI.
On the
contrary, structural issues for AI for sustainable development, such as
its guarantee of respect for human rights, the construction of AI governance
frameworks oriented towards sustainability, or the sustainability problems
generated by AI development are not analyzed. The last two elements speak of a
shallow integration between AI and the environment in SDGs national reporting.
In this
way, mentioning AI seems to be a cliché that justifies itself (we must
invest in AI because AI is important) rather than a well-thought-out decision.
It is crucial to ensure that AI integration in sustainable development public
policies is a strategic decision to obtain concrete results.
This
self-justification for the inclusion of AI is especially concerning to me, as
it creates an avenue towards emptying a real integration of AI in sustainable
development efforts. Instead, AI risks becoming part of an
"issues-to-mention" in an updated and cool checklist, which helps
explain why every country mentions AI in their speeches while failing to
understand its opportunities and risks.
It's true
that when an issue becomes "necessary to mention" it can end up
occupying a substantial space over time: this happened with environmental
issues, with gender, and possibly is happening today with "future
generations" and with "AI." Still, the speed at which the latter
moves and the extensive experience and the number of experts working in the AI
field make it urgent to integrate better the knowledge we already have on AI
for sustainable development.
VNR
preparation processes need to include meetings with AI experts in their stakeholders’
consultations, and they must do so urgently. These experts are ready to provide
valuable insights and guidance on effectively integrating AI into sustainable
development plans.
Promoting
AI development without establishing strong regulatory and oversight frameworks
or giving it a concrete direction will only increase medium and long-term
risks, creating short-range mirages of progress.
In AI
for sustainable development, talking just for the sake of talking is an
insufficiently studied risk.
2024 Voluntary National Reports analyzed:
Armenia, Austria,
Azerbaijan, Belize, Brazil, Colombia, Costa Rica, Ecuador, Equatorial Guinea, Eritrea,
Georgia, Honduras, Kenya, Lao (PDR), Mauritius, Mexico, Namibia, Nepal, Palau, Peru,
Sierra Leone, Solomon Islands, South Sudan, Spain, Uganda, Vanuatu, Yemen, Zimbabwe.
