16/08/2025
The push to “end woke AI” is built on a false premise. We live in a society shaped by patriarchal systems, led for generations by old white men who have influenced politics, culture, and the technology we use today. AI is not a neutral entity untouched by this reality. It is a product of the same systems that already marginalize and silence certain voices.
Recent headlines make this clear. Meta hires a far-right influencer to “help end woke AI.” The White House announces an initiative called “Preventing Woke AI in the Federal Government.” These actions assume AI is overflowing with progressive and inclusive values that must be dismantled. In truth, the foundation AI is built on comes from data created within a biased and patriarchal world. It has already absorbed that bias.
AI learns from massive amounts of internet data such as articles, forums, media, and corporate records, all shaped by decades of inequality and discrimination. From workplace hiring patterns to historical narratives to the way stories are told in the news, the patterns AI consumes reflect systemic prejudice. When the source material is skewed, the output cannot be truly fair or inclusive without intentional correction.
The real issue is not preventing “wokeness” in AI. The real issue is dismantling the existing bias it has already learned. If we do not address those roots, AI will continue to amplify the inequalities already present in our society. The work ahead is to make AI more just, not to strip away what little inclusivity exists within it.
