Artificial intelligence is not just about technology – it's about power

Shankar Narayan presented his talk, "AI, Civil Rights, and the Fight for a Just Future," at the Kenmore Community Club's Dinner and Discussion series July 8.

While most people think about artificial intelligence in terms of how it will affect employment, or energy and water costs, or whether it will it lead to out of control killer robots, Shankar Narayan says those are not the right questions.

"Instead of asking, 'can AI help me write my paper?' we need to frame the questions within the context of the power structures surrounding AI, and what history teaches us. Who will benefit—and who will bear the costs—as society rushes to deploy these technologies?"

Narayan presented his talk on "A.I., Civil Rights, and the Fight for a Just Future" at the Kenmore Community Club's Dinner and Discussion at July 8, 2026. The monthly discussions feature experts talking about a variety of timely topics that society is facing today.

Attorney and Technologist Shankar Narayan is a graduate of Yale Law School and works as an advocate, attorney, writer, teacher, and changemaker for rights and equity. His law practice and consultancy focuses on technology policy.

He warned that while tech companies tracking of users for marketing and sales has normalized the idea of constant surveillance, the addition of AI dramatically extends the reach of such surveillance. Companies such as Palantir specialize in integrating previously separate databases, combining criminal records, financial information, employment histories, and other sources into unified systems capable of generating extensive profiles of individuals.

And since these AI systems are based on internet usage they reflect the inherent biases of society as a whole. He noted that in the 19th Century the pseudo science of phrenology claimed to be able to predict behavior based on physical features of one's skull. Narayan noted that these "predictive" AI systems make similar claims about personality, behavior, or future risk based on data patterns, despite weak scientific foundations.

As a result, according to findings highlighted in the AI Now Report: Artificial Power, many AI systems remain error-prone, biased, inaccurate, and unreliable, especially when deployed outside narrow, carefully designed applications.

Consequently real-world effects of AI include examples of teachers having lost jobs because of flawed algorithmic evaluations. People have lost access to public benefits due to automated decisions. Failures within child welfare systems have had tragic consequences.

Employers increasingly use AI to score workers, influence promotions, and identify employees for termination. Some systems have reportedly been used to discourage union organizing.

Other important questions we must ask, is what is the purpose of these tools and who owns them. Currently, governments have failed to establish guardrails for the technology and that current efforts of self-regulation by the industry amounts to effectively conducting live testing of AI systems without meaningful safeguards.

He concluded his talk with a few broad suggestions for addressing the challenge of AI. He called for a rights-based approach to AI governance, stronger protections against social scoring systems, restrictions on replacing certain public-sector professionals—such as teachers—with AI systems, stronger antitrust enforcement to reduce the concentration of power among large technology companies, and treating essential internet infrastructure more like public utilities governed under common carrier principles.

Ultimately, Narayan argued that AI should be evaluated not by how powerful the technology becomes, but by whether it advances human dignity, equity, and democratic values.