In April 2021, the European Commission introduced the world's "first rules on AI," proposing the initial EU regulatory framework for AI. Fast forward to 2023 and the the European Parliament adopted its negotiating position on the AI Act. However, numerous countries and regions worldwide still lack any form of regulation concerning AI and the utilization of innovative technology. 

The questions surrounding AI outnumber the answers: How can we ensure that AI systems are safe, transparent, traceable, non-discriminatory, and environmentally friendly? 

As civil society organizations, acting as intermediaries between governments and communities, we are already experiencing profound effects of AI. From Hong Kong to Belarus, we have witnessed firsthand the existence and manifestation of AI technologies in the real world. Activists and frontline defenders have suffered abuses as a result of AI technologies, ranging from facial recognition to other forms of data collection and biases. 

Simultaneously, various machine learning models are aiding in detecting risks to civic space, and numerous activists and technology experts are investing in "ethical" or "responsible" AI to mitigate associated risks and support the work of civil society organizations. 

Drawing upon initiatives such as the Decolonial AI Manyfesto and the EU Artificial Intelligence Act for Fundamental Rights, this Civil Society Manifesto for AI represents a collaborative effort to advise policymakers on the risks and opportunities of AI specifically for civil society organizations and its impact on their work.  

It brings together diverse voices, including those of indigenous peoples, youth, women, and human right defenders. The manifesto acknowledges the various impacts AI is having on communities and offers recommendations on building a decolonial, fair, accessible AI that can expand rather than restrict civic space worldwide. 

The manifesto, developed by a group of over 40 organizations from 30+ countries, will be presented and its messages will be followed up at national, regional, and international levels. This collective effort aims to engage in advocacy and capacity strengthening activities to ensure the document's impact. 

We begin with the assumption of researchers and activists, who argue that AI is a product of historical data representing inequities and inequalities. If we utilize current AI systems as a basis for decision-making regarding the future, we will only perpetuate existing inequities. 

We also acknowledge that the discourse surrounding AI, like the technology itself, has predominantly been shaped by “Western, whiteness, and wealth”. Our objective is to not only explore, challenge, disrupt, and reimagine the underlying assumptions within this discourse but also to broaden the discussion to incorporate communities beyond the traditional "experts."  

We recognize that the lives, work, and future of various groups, such as young people, those impacted by the environmental consequences of AI, and individuals affected by its racial and gender biases, will be significantly influenced. In the subsequent pages, they will present their unique perspectives on AI, offering their own interpretations and understandings of the concept, policies and practices. 

We are currently at a critical juncture where most countries lack a comprehensive AI policy or regulatory framework. However, several initiatives have been undertaken to promote safer AI adoption and address potential risks associated with the technology. 

The sudden reliance on AI and other digital technologies due to the COVID-19 pandemic has introduced new vulnerabilities, allowing citizens' data to be commercialized and even weaponized. Consequently, issues such as algorithmic bias, privacy breaches, lack of transparency, and the overall complexity of understanding AI interactions require policy considerations. It is crucial to inform AI policies that approach the extent of reliance on AI systems critically, particularly in public sectors.  

Ultimately, because of its strategic role and responsibilities, civil society possesses the potential to steer AI in a positive direction. By promoting accountability and elevating the voices of historically marginalized communities who have been overlooked, we believe that this collective manifesto will contribute to achieving these goals. 
 

How the Civil Society Manifesto for AI started and who is involved 

The process of crafting the Civil Society Manifesto for AI began at the Digital Rights Asia-Pacific 2023 Assembly (DRAPAC), held in Chiang Mai in May 2023. This collaborative endeavor was initiated through a roundtable discussion organized by Forus, a global civil society network, in collaboration with the Asia Development Alliance, ECNL, the Asia Pacific Forum on Women, Law and Development (APWLD), the Asia Indigenous Peoples Pact (AIPP), and the Asia Center. Over 50 civil society organisations gathered online and offline to explore the risks and opportunities associated with AI and its applications. 

Subsequently, a virtual global workshop was conducted to delve deeper into these topics with diverse testimonials from civil society organizations, including Paradigm Initiative, the Uganda National NGO Forum, Red Unitas from Bolivia, the Asia Indigenous Peoples Pact (AIPP) and initiatives such as Machine Learning for Peace project from the University of Pennsylvania. The workshop aimed to foster collective discussions and share personal experiences regarding AI's impact on the operations and working environment of civil society organizations. The workshop was followed by an interactive session as part of the Tech for Democracy Advisory Group meeting. 

The organizations and activists involved in these initiatives are forming the constituting body responsible for developing the Civil Society Manifesto for AI, a collaborative writing process spanning three months.