Page 41 - AI for Good - Impact Report
P. 41
AI for Good
Sustainable Development Goal 1: No Poverty
End poverty in all its forms everywhere
SDG 1 is facing a critical situation, with none of its 7 targets
on track as of 2024. The UN warns that if current trends
149
persist, an estimated 575 million people will still be living in
extreme poverty by 2030, greatly impacting their quality of
150
life and well-being. The UN further highlights that global
efforts to eradicate extreme poverty have suffered setbacks
due to the COVID-19 pandemic and other major shocks,
leading to the first increase in extreme poverty in decades
and reversing global progress by three years. 151
AI and SDG 1 Impact
According to a study on the impact of AI on
AI solutions and technologies can have various impacts on SDG 1. SDG 1 could act as an (positive) enabler for
First, the technology may indirectly contribute to SDG 1’s advance- 100% of the targets and act as an inhibitor
(negative) for 86% of the targets. (Nature
ment by strengthening other SDGs such as SDG 9 or SDG 8. 152 153 Communications, 2020)
By enabling research and innovation, the benefits generated could
trickle down to SDG 1 via the creation of new products or services
that are more affordable or better suited to the needs of the most Use case 1
vulnerable communities. For instance, using AI to reduce the costs Using AI to make the process faster for
associated with agricultural practices (e.g., minimizing the use of micro-finance loans and to provide access
fertilizers) could enable communities to enhance their quality of to financial services to communities that
traditionally have been underserved.
life. 154 This indirect influence represents the most significant posi-
tive impact of AI on SDG 1. Moreover, government support for AI in
innovation and economic growth could indirectly lead to improve-
ments in SDG 1.
Specific use cases linked to each target of SDG 1 can yield addi-
tional impact. For example, AI can enhance the efficiency of the
financial sector, thereby increasing accessibility for the 1.7 billion
adults lacking access to financial services. 155 However, the number
of AI use cases for SDG 1 is less than other SDGs, reducing the link
collaborative efforts between the technology and the Goal. For
instance, based on two UN reports, there are only 2 use cases out of Use case 2
40 in the AI for Good: Innovate for Impact report, 156 and around 70 Improving climate forecasting to better
use cases out of 408 in the UN Activities on AI report. 157 prepare communities for extreme weather
events and reduce exposure to climate risks
While AI can generate positive impacts for SDG 1, the potential risks as aimed in target 5 of SDG 1.
associated with the technology for this Goal must be considered. AI
could widen inequality between countries, 158 and individuals. 159 160
Ownership of AI solutions could further create monopolies, leading
to a further concentration of wealth and power without equitable
compensation for the content providers. Additionally, investments
in AI and its infrastructures, such as robots in agriculture, can be
costly and may hinder access to technology for the poorest commu-
nities, further widening the wealth divide. 161 To assure fairness and link
value for all, governments should account for new value-sharing
models in their legislation. Use case 3
Providing access to overlooked commu-
Key Considerations for Stakeholders nities to market solutions to provide them
with a new revenue stream and higher
• Technology access: A significant constraint in AI adoption is financial resilience.
the cost of associated access to market, in terms of AI hard-
ware and software. For instance, not everyone can afford
robots to assist in their operations in agriculture. 162 Imple-
menting a sharing model or subsidizing hardware costs
could mitigate this risk.
• Ownership sharing: To mitigate the risk of technology
monopolies, it is important to explore new business models
that distribute value differently, 163 aiming to minimize link
inequality growth and fairly reward all contributors.
31