This past week, India hosted the AI Impact Summit—the world’s largest-ever global gathering of world leaders and CEOs of AI companies. This summit, while the largest, is not the first. Such an event has become part of an annual series aimed at charting the future for AI, beginning in Britain in 2023, followed by North Korea and France. As the first country in the Global South to host this annual event, the discussions in India primarily focused on the economic and developmental prospects of AI, with many speakers presenting it as a potential remedy to real-world problems. At the same time, however, there was notably less discourse foregrounding regulation and safety concerns. Instead, the sentiment expressed by government leaders and tech giants largely emphasized the need for innovation and investment, with India’s prime minister, Narendra Modi, proclaiming, “We must give AI an open sky.” In reality, the unpleasant truth is that this techno-solutionist rhetoric obscures the human rights violations AI systems have worked to further, particularly for marginalized communities. I argue that we must not blindly support the growth of the AI industry for the sake of advancing innovation and competition, but rather, our conversations must focus attention on promoting binding rights protections that will be effective in safeguarding people’s rights and welfare.

For all the talk of development and collective solutions, the summit appeared to serve as a stage for India to advance its long-desired mission of being a leader in AI with the goal of stimulating economic growth and serving as the primary export hub for countries in the Global South. Despite efforts to use technology as a domestic and foreign policy tool, India has faced several challenges, especially considering it does not have access to homegrown tech companies such as OpenAI and Nvidia, as the U.S., or the rare earth minerals necessary to build high-powered microprocessors that fuel AI, as China does. What India does possess, though, is access to a diverse talent pool and partnerships with the U.S., European Union, and Britain, which have supported its ambitions. During the summit, these partnerships deepened, as multiple AI companies announced deals with Indian companies. In doing so, India is actively attempting to carve out a space for itself amid the AI arms race currently taking place between the U.S. and China.

Throughout the week, world leaders praised India for its strides toward technological progress, while strikingly disregarding the human rights abuses such technologies have facilitated, both within the country and internationally. To take one example, no word was mentioned about the harmful impacts of algorithmic systems such as India’s Samagra Vedika, which has been used by state officials to determine the eligibility of welfare recipients by combining data from individuals across various government databases. According to an Amnesty International investigation, this system has left many without basic necessities due to misidentifications that wrongly classify vulnerable individuals and families as disqualified for social services, with officials often accepting the algorithm’s word over citizens.

Additionally, it is important to note that these risks extend far beyond India. Oppressive uses of AI, whether through facial recognition technologies, automated social credit systems, or predictive policing, have been documented in the U.S., Israel, China, and Russia, among others. Although advocates of these systems claim that AI-enabled systems will authorize quicker and more efficient enforcement decisions, in application, the assessments generated by these technologies consistently produce false positives, given that they are created by humans and inevitably reflect prevalent societal judgments and biases.

As human lives continue to be reduced to pieces of data, we must recognize that speaking of AI as an egalitarian, democratic tool capable of providing developmental opportunities will fail to address the mounting harms caused by its deployment. Future summits must move away from the performativity of rights-based language and begin taking actionable steps toward binding guardrails by prioritizing the voices of affected communities as well as civil society members who are dedicated to ensuring that AI systems are compatible with ethical standards. In the meantime, it will be necessary that all states administer comprehensive human rights evaluations of these technologies, starting with their design all the way through deployment, in order to best uphold transparency, fairness, and pathways to remedy.