While the energy- and emissions-intensive nature of AI has been widely publicized, the intersection of AI and climate risk goes beyond just the GHG footprint – as tech giants have realized at their peril. In December 2024, Microsoft shareholders voted on a resolution to mandate that the firm report on risks associated with providing AI and other advanced technologies that support new oil and gas development. The resolution maintained that these technologies undermine Microsoft’s climate mitigation targets and expose the firm to reputational, operational and competitive risks.
Ultimately, the resolution did not pass – only 10% of shareholders voted in favour – but the risks will not disappear. After about three years of internal lobbying, several Microsoft employees have resigned in protest of these new oil and gas contracts; at least two left in early 2024 and published their motivations for doing so. Employees are speaking out, sounding the alarm that the tech giant does not report on emissions from fossil fuel production enabled by its technology. Microsoft may face greenwashing accusations.
While it may be Microsoft under the spotlight at the moment, it is not alone in offering advanced technologies designed for oil and gas production. In 2020, Greenpeace published a report highlighting ties between the oil and gas industry and big tech – specifically, Amazon, Google and Microsoft. This produced significant results – and quickly. Within days of the report’s publication, Google pledged to stop offering customized AI for oil and gas extraction; Amazon and Microsoft made no such commitment.
Being seen as less green than competitors comes with reputational risks and the potential to lose market share. Many organizations consider sustainability when choosing suppliers, even discontinuing existing supplier relationships due to their carbon emissions performance. Ties with the fossil fuel industry make organizations less attractive to firms worried about supply chain and third-party risk. While tech firms are keen to highlight AI’s ability to do good, developing tools for the fossil fuel industry will compromise their reputations.
Climate-related risks interact with many other types of risk. With demands for transparent climate risk reporting and non-financial risk management ramping up, organizations should expand the boundaries of analysis. Established frameworks are unlikely to provide the scope to measure a product’s indirect emissions, or other far-reaching impacts of a business’s activities – but stakeholders are paying attention. Organizations exist within a complex and interconnected world. Diverse risks are related. If decision-makers neglect this fact, it is likely to lead to poor risk management and, ultimately, hurt the bottom line.