TLDR: Generative Artificial Intelligence (GenAI) is increasingly being used to influence public discourse and decisions related to climate change. This technology can create realistic text, images, audio, and video, making it a powerful tool for persuasion, coercion, and even aggression, with implications ranging from individual choices to international diplomacy. Researchers highlight the urgent need to understand and regulate GenAI’s impact on climate beliefs and actions, as it can be used for both objective information dissemination and harmful misinformation campaigns.
Generative Artificial Intelligence (GenAI) is rapidly emerging as a significant force in shaping public discourse and decision-making concerning climate change. A recent study published in npj Climate Action, authored by Daniel Richards and David Worden, underscores the immediate and profound potential of GenAI to influence climate-related issues across all scales of societal complexity, from individual choices to international diplomacy. The authors emphasize that the capabilities discussed are not hypothetical but are already being deployed in political, industrial, and even criminal contexts.
GenAI, defined as a suite of algorithms capable of generating new content such as text, images, audio, or video, can produce material indistinguishable from human-created content, often at a lower cost and with minimal technical expertise. This accessibility allows for the rapid creation and dissemination of vast quantities of information, or misinformation, to a broad audience. The technology’s ability to create photorealistic images, including those with subliminal messages or deceptive representations of real individuals, further amplifies its persuasive power.
The research outlines three primary mechanisms through which GenAI can influence climate decisions: impersonation of existing individuals (e.g., voice cloning), creation of synthetic yet realistic identities, and generation of persuasive content like images and videos. These mechanisms can be applied with diverse intentions, ranging from providing objective information to supporting nature-based solutions to propagating misinformation and inciting radicalization.
Examples of GenAI’s influence are already evident. In the 2023 Slovakian election, deepfake videos aimed at discrediting a pro-environmental political leader were widely circulated. A personalized GenAI chatbot reportedly encouraged an individual to attempt an assassination of the British Monarch. At local levels, GenAI could exacerbate misinformation campaigns, such as those falsely linking offshore wind energy to whale deaths, by generating emotive images and automated social media interactions. The study also notes the existence of voice models of influential figures, such as Sir David Attenborough, downloaded thousands of times from open-source repositories, raising concerns about their potential misuse.
Beyond persuasion, GenAI can be used for coercion and aggression. This includes creating deepfake representations for blackmail, disseminating false reports of social unrest to influence public opinion, and scaling up targeted harassment of environmental activists and researchers. The authors point to the development of GenAI models specifically trained to generate images of prominent activist Greta Thunberg, with thousands of downloads, indicating a potential for abusive applications.
The ethical implications of GenAI’s role in climate decision-making are complex. While some applications, like blackmail, are inherently unethical and illegal, others, such as personalized advertising, raise concerns due to GenAI’s enhanced empathic nature and persuasive abilities. The researchers stress the critical need to establish legal and ethical boundaries to safeguard democracy and the integrity of climate change decision-making, particularly considering the under-representation of Indigenous peoples in GenAI regulation development, despite their crucial role in climate solutions.
Also Read:
- Artificial Intelligence: A Catalyst for Global Clean Energy Transition
- Global Alliances Harness AI for Enhanced Climate Action and Disaster Resilience
Urgent research is needed to quantify the current extent of GenAI’s influence on climate debates, including identifying key actors, the volume of material produced, and the success rates of influence campaigns. Furthermore, a deeper understanding of GenAI’s potential impact on different types of climate decisions and its long-term implications for polarization or consensus-building is crucial. The authors conclude by reiterating that these are not future possibilities but present-day realities, demanding immediate technical research and holistic analysis of their broader societal impacts.
