Page 54 - AI for Good - Impact Report

AI for Good

Sustainable Development Goal 13: Climate

Action

Take urgent action to combat climate change and it's
impacts

SDG 13 shows limited progress, with 2 out of 5 targets
advancing moderately 340 This lack of progress could lead
to 2.5°C degree warming by 2 100, 341 posing a direct threat
to all other SDGs. For example, climate change cost the
economy an average of US$803 billion between 2019 and
2020, impacting SDG 8, and increasing mortality rates in
vulnerable regions, affecting progress towards SDG 3. 342

AI and SDG 13 Impact
According to a study on the impact of AI on
The relationship between AI and SDG 13 is well-documented, with numer- SDG 13, AI could act as an (positive) enabler
ous use cases highlighted in various UN repositories: 7 use cases out of for 80% of the targets and act as an inhibitor
40 in AI for Good: Innovate for Impact, 343 and around 110 use cases out of (negative) for 20% of the targets. (Nature
Communications, 2020)
408 in the UN Activities on AI. 344 Similar, to other environmental SDGs, the
synergy between AI and SDG 13 is a paradox. On one side, numerous use Use case 1
cases can enhance climate actions by using AI, while on the other side, the Using AI to optimize the CO2 emissions of
energy usage and increase in consumption behaviors from AI put the entire organizations from transport, distribution
relationship at risk. 345 Interesting use cases from AI include the optimiza- and logistics.
tion of logistics such as freight roads to minimize CO₂ emissions, where the
road used is for example the least carbon-intensive. 346 This is a valuable
reduction, as transport accounts for around one-fifth of global CO₂ emis-
sions. 347 Additionally, AI can be used to drive CO₂ measurement and can give
additional visibility to the causes and effects of climate change to govern-
ments. 348 It can also help governments better monitor their climate impact
and make informed decisions around it. 349 Moreover, AI can also be used to
provide improved forecasting abilities on weather events, to help govern-
ments and organizations better prepare for adverse climate events, and also link
better prepare the regions to support such catastrophes. 350 Additional AI Use case 2
use cases for the climate include mapping melting behaviors of icebergs, Leveraging AI to assess the new frontiers of
helping communities at risk to better mitigate climate impact or supporting climate science and provide a better under-
organizations in finding pathways to decarbonize. 351 AI can give the right standing of our climate and climate change.
tools to governments to better predict and plan the challenges that climate
change will generate.
The rise of AI use cases, however, can also be problematic for SDG 13 as AI
consumes important energy, which does not only originate from renewable
energy, and lead to the creation of 0.01% of the GHG emissions currently. 352
The question arises with the socialization of AI to see if this number is going
to rise, as the demand for AI by 30% - 40% annually. Developing new efficient
models could be a pathway to minimize any growth impact of AI. Another
problem could regard the rise of new use cases such as marketing-driven
cases that push for consumerism, and thus for the purchase of new prod- link
ucts and services that would generate GHG emissions to be produced and Use case 3
managed. 353
Implementing AI forecasting solutions for
weather events to better prepare communi-
Key Considerations for Stakeholders ties for adverse weather events and the risks
associated.
• Impact assessment: Aligning AI use case development and incentives
with OECD AI principles to maximize sustainable value creation. 354 The
objective is to prioritize governmental tools for AI use cases related to
the SDGs.
• Technology improvement: Reducing energy consumption is imperative
to support the development of SDG 13, hence technologies with less
energy requirements should be prioritized. 355

link

44