Whereas: Amazon’s Responsible AI Approach is guided by eight priorities, which include “fairness” (evaluating AI’s impacts on different groups), “privacy and security” (appropriately obtaining and using data), “safety” (preventing harmful system output and misuse), and “transparency” (enabling stakeholders to make informed decisions about their engagement with AI).1 Despite this approach, Amazon continues to sell to and maintain contracts with entities engaged in rights-violating applications of its AI and related technologies, suggesting misalignment between policies and practice. Such misalignment presents material legal, reputational, regulatory, and litigation risks to Amazon and its investors. For example, Amazon’s cloud, AWS, is the world’s most broadly adopted cloud and provides cloud computing, artificial intelligence, and data storage.2 Its $1.2 billion cloud computing contract with Israel - Project Nimbus - has been used by Israel in its attacks on Palestinians, actions prominent organizations have classified as genocide.3 Although the contract’s details are not publicly available, a 2024 investigation found Israel had used AWS to store masses of military and surveillance data, and in some cases, to aid in airstrikes that killed many civilians.4 An October 2025 investigation revealed the contract prohibits AWS from suspending, withdrawing, or restricting use of its technologies - even if Israel breaches AWS’ terms of service - and requires AWS to violate court-imposed gag orders.5 Amazon’s AI and related technologies also enable the US’ expansive immigration crackdown, in which US Department of Homeland Security (DHS) agencies have been accused of arbitrary detentions, silencing free speech, and violating rights to privacy, nondiscrimination, asylum protections, due process, and other human rights.6 AWS hosts many DHS databases and systems used to track, monitor, and deport immigrants.7 AWS will host DHS’ Homeland Advanced Recognition Technology (HART) system, which will rely on AI to store and process information about immigrants to facilitate detention and deportation.8 1 https://www.aboutamazon.com/what-we-do/artificial-intelligence-ai/responsible-ai 2 https://aws.amazon.com/what-is-aws/?nc1=f_cc 3 https://www.un.org/unispal/wp-content/uploads/2025/09/a-hrc-60-crp-3.pdf 4 https://www.972mag.com/cloud-israeli-army-gaza-amazon-google-microsoft/ 5 https://www.theguardian.com/us-news/2025/oct/29/google-amazon-israel-contract-secret-code 6 https://rfkhumanrights.org/wp-content/uploads/2025/04/FINAL-UPR-Imm.-CoalitionSubmission_4.7.25.pdf; https://www.youtube.com/watch?v=5zJpvgzxnwg&t=6s 7 https://mijente.net/wp-content/uploads/2018/10/WHO%E2%80%99S-BEHIND-ICE_-The-Tech-and-DataCompanies-Fueling-Deportations-_v1.pdf 8 https://surveillanceresistancelab.org/wp-content/uploads/2023/01/HART-Attack-2022.pdf Amazon’s Ring, which produces doorbell cameras and has extensive partnerships with police departments,9 recently announced it will enable facial recognition,10 drawing sharp criticism for creating new privacy and civil liberties risks.11 Additionally, Ring announced a partnership with Flock, a network of AI-powered cameras whose data has been utilized by US Immigration and Customs Enforcement (ICE) in its immigration enforcement,12 further increasing the risk that Amazon’s AI will be misused to violate human rights. Moreover, because of the lack of transparency about the specific ways Amazon’s AI and related technologies are being used, investors cannot be assured there is alignment with its commitments.