In the context of AI resistance refers to a broad range of behaviors and discourses that might be overt or covert, organized or diffuse, individual or collective, positional or reformist. Resistance to AI can be viewed as a type of agency used within current power structures, drawing on ideas from critical theory and science and technology studies. The sociotechnical arrangements and imbalances that both shape and are shaped by the development and use of AI are the object of opposition in this framing rather than technology per see.
AI resistance: Who rejects AI and why?
a contaminated dataset. Hollywood was put on hold for 148 days due to a writers’ strike. demonstrations against data centers on the streets. A growing worldwide backlash against artificial intelligence is the driving force behind each of these actions. Based on Can Sims and apse Gizem Yasar’s recent report, “From Rejection to Regulation:
This essay, “Mapping the Landscape of AI Resistance,” examines the ways in which academics, artists, workers, and activists are against the creation, application, and management of AI systems. It examines the motivations behind AI resistance and presents a research program that views these actions as essential contributions to democratic AI governance rather than as roadblocks.
A profound sociotechnical revolution is being sparked by artificial intelligence, which is changing not just our technology infrastructures but also the institutions that govern society. During this transition, important questions come up: Who decides the course of this transformation and the future we choose to create? In the conversation, who is not heard? Do we have the agency and obligation to question and change the increasingly powerful algorithms that are being used, or are we merely passive observers?
AI resistance: From learning machines to weaves
It is helpful to keep in mind that opposition to new technology is nothing new in order to put today’s opposition to AI in perspective. From the Luddites of the 19th century who opposed textile technology due to labor displacement to the contemporary discussions on digital monitoring and algorithmic prejudice, technological paradigm shifts have continuously sparked public concern and opposition. The public’s responses to artificial intelligence’s emergence as a significant trans-formative force have been alternating between optimism and anxiety. On the one hand, governments and private companies are investing at previously unheard-of levels in the development of AI; on the other hand, an increasing amount of “AI resistance” raises basic concerns about the design, production, deployment, and governance of these technologies. However, what precisely is resistance to AI?
Examples of resistance to AI
In the digital world, historical means of resistance find new equivalents. For example: The term “sabotage” was used to describe intentional actions that interfered with industrial machinery in early 20th-century France. It is believed that workers threw their wooden shoes, or “sabots,” into machines to halt production. A similar strategy can be found in “data poisoning” in today’s digital environment. This phenomena involves artists and other creators subtly altering their work to deceive AI models, for as by subtly altering text or graphics, so that when AI systems employ these resources for training, the resulting models are degraded or misled.
This is not merely a technical assault. It stops the unauthorized use of creative work in AI training by deliberately degrading the algorithms and transforming data poisoning from a technical flaw into a form of resistance.
In addition to employing these tactics, workers in the creative industries have used more conventional modes of resistance, such as going on strike, to defend their rights and have an impact on how new technology are incorporated into their jobs. The 2023 Hollywood writers’ strike, in which the Writers Guild of America demanded contractual provisions against the use of AI to compose or rewrite scripts or to use AI-generated content as source content, is a well-known example.
From demonstrations to legislation
We documented many examples of AI resistance in the paper, including demonstrations against data centers’ negative environmental effects, opposition from large tech workers to AI’s military uses, and public outcry over the UK’s A-level grading scandal. We looked at six major areas where this kind of resistance has been very prevalent, but this was not meant to be comprehensive:
(i) the creative industries
(ii) border security and migration
(iii) medical artificial intelligence
(iv) higher education
(v) the defense and security sectors and
(vi) environmental activism.
As a result, we highlighted important players in the fight against AI, paying special attention to the contribution civil society plays in organizing public opposition. The report also examines how some types of resistance have been enacted by governments. The EU AI Act, which forbids some AI systems, such as intentionally manipulative AI techniques, is one example.
Why do people resist AI?
- (i) First, there are socioeconomic issues, like as in the creative sectors, where the 2023 Writers Guild of America strike targeted AI’s ability to displace human labor.
- (ii) Second, when AI systems are biased or opaque, such migration risk-assessment tools that can unjustly affect people’s future decisions, ethical problems emerge.
- (iii) Third, safety issues are a concern, particularly in the healthcare industry, where medical professionals have voiced concerns due to faulty AI diagnostic results.
- (iv) Fourth, the use of AI for widespread social manipulation poses a threat to democracy and sovereignty.
- (v) Lastly, there is the environmental impact: research demonstrating the substantial carbon footprint of training large AI models has been highlighted by climate-focused NGOs.
Outlining a research road-map on AI resistance
Research on AI resistance needs to be advanced in order to shed light on these ubiquitous societal challenges that are sometimes ignored in technology and policy-oriented discourse. These voices of opposition should not be dismissed as mere dissent; rather, they are crucial first steps toward political systems that better reflect democratic values and the common good. The scope and depth of civic engagement can be seen by mapping the activities and motivations behind AI resistance. Participation in the development, application, and regulation of AI technology is being demanded by workers, artists, physicians, educators, activists, and academics. We hope to create space for more diverse and socially conscious discussions about technological futures by acknowledging these interventions as significant contributions rather than barriers.
