Introduction and Methodology
INTRODUCTION
Willingly or unwillingly, billions of people around the world have come to interact with or be affected by Artificial Intelligence.
Willingly or unwillingly, billions of people around the world have come to interact with or be affected by Artificial Intelligence (AI) in some way or another on a daily basis through their use of digital services and platforms. Content recommendation systems on social media platforms, 15 streaming services, online shopping platforms and search engines shape the content people see and consume. 16 Content moderation algorithms are used to detect undesirable content and remove it. 17 Algorithms deployed by digital assistants recognise speech to help users perform an action, such as placing an online order or playing a song. 18 Smartphones, messaging apps, and email use AI to predict what users are typing and help them write messages faster.
Rapid advancements and wider adoption in various industries, including banking, healthcare, security and defence, welfare allocation and human resources, has not only made AI more ubiquitous but has also given rise to far-reaching ethical, societal, economic and political implications. 19 In particular, military deployments in autonomous weapons are a threat to the right to life and violate international humanitarian law. 20 Additionally, a few powerful tech giants are infringing on the right to privacy, stealing copyrighted work, and exploiting data workers, mostly in the Global South, to train and develop their models. 21 Further, the unchecked AI advancements are exacerbating the climate crisis due to the excessive use of water resources. 22 Of concern, public sector deployment of AI systems, for instance, in border and migration management, 23 policing 24 and welfare allocation, 25 contribute to the discrimination of excluded groups and amplify inequalities.
Civil society’s position in relation to AI deployers, developers, and regulators is unique and shaped by power asymmetries
Civil society, as any other sector, is not immune to AI’s disruptive implications and impacts; however, its position in relation to AI deployers, developers, and regulators is unique and shaped by power asymmetries. First, civil society organisations (CSOs) face AI-enabled threats from state and non-state actors that negatively interfere with their human rights work. These include facial recognition systems used for surveillance of activists and protesters, 26 content moderation algorithms that suppress dissent, and AI-generated disinformation campaigns that erode information integrity and public trust. 27 Second, as CSOs and activists experiment with AI and increasingly adopt it in their work, such as in fact-checking, operations, interactions with audiences and in creating content, they must grapple with the ethical questions and potential risks to human rights. 28 These risks include amplifying biases and the spread of misinformation, over-reliance on corporate models from the Global North, particularly the U.S., and compromising personal and sensitive information. 29 Without adequate safeguards, CSOs risk perpetuating the same harms they seek to address. Third, civil society is a key actor in advocating for the respect and protection of human rights in AI deployment and governance. 30 Yet, they face multiple challenges, including interference with their work, civic space under pressure, funding challenges 31 and obstacles in participating in AI governance discussions, which are dominated by a few powerful governmental and private sector actors. 32
This research, therefore, aims to explore these three power asymmetries pertaining to AI and civil society. It aims to provide a better understanding of the risks of AI to CSOs around the world and the challenges they face in the deployment of the AI and in AI advocacy and accountability efforts.
The research is guided by the following questions:
- What risks does the deployment of AI by state and non-state actors pose to civil society?
- How do CSOs deploy AI in their work? What challenges and ethical questions arise from their deployment of AI and how do they attempt to address them?
- What challenges does civil society face in engaging in AI accountability and advocacy efforts?
The first chapter examines the risks of AI for civil society groups and their work. The second chapter of the report sheds light on the extent of AI adoption by civil society, including how they deploy it and navigate risks to human rights and ethical questions. The third chapter looks at how CSOs have been advocating for human rights safeguards in the deployment and governance of AI systems and the different challenges they face in this regard. The report concludes with recommendations to improve support to civil society in their adoption of AI in ways that align with human rights standards and advance a rights-based approach to the adoption and regulation of AI.
METHODOLOGY
The methodology consisted of desk research, a survey of CIVICUS members, interviews and an online focus group.
Abour this research
Literature review
293 survey responses
7 expert interviews
1 focus group (8 CSO reps)
1.5 hours of discussion
First, a comprehensive literature review was conducted to identify trends and key gaps in knowledge. The review of academic studies, AI policy documents, civil society reports, and credible media reports focused on identifying 1) risks posed by AI to civil society—such as state surveillance, algorithmic content moderation, and disinformation; 2) risks emerging from civil society’s adoption of AI tools without adequate safeguards; 3) obstacles and challenges that prevent CSOs from deploying AI or engaging in the field of AI accountability; and 4) a review of key policy trends in AI regulation globally.
Based on the literature review and the research questions, a survey was developed and disseminated among CSOs that are members of the CIVICUS network. The survey questions addressed the following:
- AI systems and applications that pose risks to CSOs and how they impact human rights and their work.
- To what extent CSOs are deploying AI in their work and the challenges they face in AI deployment, and what (if any safeguards) do they have in place to address and mitigate risks associated with such adoption such as bias, privacy, technological dependence on U.S. models, etc.
- To what extent CSOs aim to shape how AI is developed, deployed, and regulated and the obstacles they face in advocating for human-rights based deployment and governance of AI (such as lack of funding and resources, algorithmic opacity, lack of opportunities to develop required knowledge and skills, etc.).
The survey was disseminated among CIVICUS members and 293 responses were received. The interviews were conducted with seven experts working at the intersection of human rights and AI and technology to help gain more in-depth insights and address any gaps in data.
Subsequently, a 1.5-hour focus group discussion (FGD) was conducted remotely with eight representatives of CSOs that have experience with deploying AI to discuss their experience and how they approach ethical questions and govern AI use internally. This FGD helped inform the development of an introductory toolkit on how CSOs can deploy AI in a way that aligns with human rights standards.
15 Adib Bin Rashid, MD Ashfakul Karim Kausik, AI revolutionizing industries worldwide: A comprehensive overview of its diverse applications, Hybrid Advances, Volume 7, 2024, 100277, ISSN 2773-207X, https://doi.org/10.1016/j.hybadv.2024.100277. (https://www.sciencedirect.com/science/article/pii/S2773207X24001386).
16 Ibid.
17 Gorwa, R., Binns, R., & Katzenbach, C. (2020). Algorithmic content moderation: Technical and political challenges in the automation of platform governance. Big Data & Society, 7(1). https://doi.org/10.1177/2053951719897945
18 ISO. “Unravelling the secrets of natural language processing”. https://www.iso.org/artificial-intelligence/natural-language-processing (accessed 30 April 2026).
19 Adib Bin Rashid, MD Ashfakul Karim Kausik, AI revolutionizing industries worldwide: A comprehensive overview of its diverse applications, Hybrid Advances, Volume 7, 2024, 100277, ISSN 2773-207X, https://doi.org/10.1016/j.hybadv.2024.100277. (https://www.sciencedirect.com/science/article/pii/S2773207X24001386).
20 CIVICUS. “‘The lesson from Gaza is clear: when AI-powered machines control who lives, human rights die’”. 16 July 2025. https://lens.civicus.org/interview/the-lesson-from-gaza-is-clear-when-ai-powered-machines-control-who-lives-human-rights-die/ (accessed 3 May 2026).
21 Hao, K. (2025). Empire of AI: Inside the reckless race for total domination. Penguin.
22 Ibid.
23 Amnesty International. “Global: New technology and AI used at borders increases inequalities and undermines human rights of migrants”. 21 May 2024. https://www.amnesty.org/en/latest/news/2024/05/global-new-technology-and-ai-used-at-borders-increases-inequalities-and-undermines-human-rights-of-migrants/ (accessed 3 May 2026).
24 Lungu, M. “Predictive policing AI is on the rise − making it accountable to the public could curb its harmful effects”. The Conversation. 6 May 2025. https://theconversation.com/predictive-policing-ai-is-on-the-rise-making-it-accountable-to-the-public-could-curb-its-harmful-effects-254185 (accessed 30 April 2026).
25 European Artificial Intelligence & Society Fund. “AI & Social Welfare Funders’ Briefing”. September 2025. https://europeanaifund.org/wp-content/uploads/2025/09/EAISF-AI-Social-welfare_Funders-Briefing.pdf (accessed 3 May 2026).
26 CIVICUS. “Facial recognition: the latest weapon against civil society”. 23 May 2025. https://lens.civicus.org/facial-recognition-the-latest-weapon-against-civil-society/” (accessed 3 May 2026).
27 CIVICUS. “CHINA: ‘The state is using generative AI to engineer reality through informational gaslighting’”. 12 March 2026. https://lens.civicus.org/interview/china-the-state-is-using-generative-ai-to-engineer-reality-through-informational-gaslighting/ (accessed 3 May 2026).
28 Asia Indigenous Peoples Pact Foundation. “Civil Society Manifestor for Ethical AI”. 2023. https://aippnet.org/wp-content/uploads/2023/12/civil-society-manifesto-for-ethical-ai.pdf (accessed 3 May 2026).
29 Digital Justice Network. “Generative AI guide for civil society”. Association for Progressive Communications. 6 January 2026. https://www.apc.org/en/pubs/generative-ai-guide-civil-society (accessed 3 May 2026).
30 CIVICUS. “‘Civil society is pushing for transparent, rights-centred AI governance’”. 18 March 2025. https://lens.civicus.org/interview/civil-society-is-pushing-for-transparent-rights-centred-ai-governance/(accessed 1 May 2026)
31 Human Rights Funders Network. “Funding at a Crossroads: Foreign Aid Cuts And Implications For Global Human Rights.” September 2025. Https://Www.Hrfn.Org/Wp-Content/Uploads/2025/09/Funding-at-a-Crossroads-HRFN-Sept-2025.pdf
32 Hickok, E., Mohan, S., Pielemeier, J. and Kakkar, J.M. “Multistakeholder Promises and Power Gaps in Global AI Summits”. 17 March 2026. Tech Policy Press. https://www.techpolicy.press/multistakeholder-promises-and-power-gaps-in-global-ai-summits/ (accessed 3 May 2026).