Recent criticism has targeted three AI-powered search engines from tech giants Google and Microsoft, as well as Perplexity, promoting unquestionably flawed and grossly discredited race science.

Their messages have unfortunately amplified a dark story: the supposed superiority of white genetic stock, according to the alarm generated by experts in the field.

The rise of racial science

WIRED discovered that Patrik Hermansso, a researcher at Hope Not Hate, a U.K.-based anti-racist organization, was continuing his investigation, testing an obscure data set claiming that IQ scores prove support for white supremacy.

Hermansson decided to test the Human Diversity Foundation, which he said was funded by Andrew Conru, a billionaire American tech mogul who co-founded Adult Friend Finder.

Founded in 2022, this group is considered a continuation of the Pioneer Fund, established by Nazi apologists in 1937 to promote “racial betterment.”

Creepy web searches

During her research, Hermansson Googled IQ scores by country. He was shocked to discover that the feature his computer turned on by default for Google’s Overview AI tool listed supposedly factual IQ scores for countries like Pakistan, Sierra Leone and Kenya.

For example, he received incredibly high, Pakistan-specific IQ scores – information taken directly from the discredited work of Richard Lynn. Lynn, a former professor at Ulster University and chairman of the Pioneer Fund, defined the racial science movement until his death in 2023.

AI-approved flawed research

The implications of AI search engines propagating Lynn’s controversial research are very disturbing.

Experts say this could have the effect of pushing people to become radicalized. Lynn’s work has been exploited and used by many white supremacists and far-right hate extremists for so long to justify their ideologies.

Recently, Lynn’s studies were cited by the 2022 Buffalo mass shooter, highlighting the real danger of these narratives.

AI tools have downsides

Google’s AI previews, launched a few months ago to respond to user queries, have drawn widespread criticism. When Hermansson asked about national IQs, the AI ​​did not rely on questionable sources for its answers.

After complaints about the issue, Google acknowledged its shortcomings and promised the tool would improve.

“We have safeguards and policies in place to protect against low-quality responses, and when we find overviews that do not align with our policies, we quickly take action against them,” he told WIRED Ned Adriance, a Google spokesperson.

Other AI platforms bear responsibility

Based on WIRED’s findings, Microsoft’s Copilot and Perplexity also repeated Lynn’s poor numbers when asked about her IQ scores. Bing even repeated false claims from a disingenuous source about Pakistan’s average IQ, while Perplexity bluntly cited Lynn’s work. These findings suggest that the spread of discredited race science is not limited to any given platform.

The need for better regulation

According to Rutherford and Rebecca Sear, director of the Center for Culture and Evolution at Brunel University London, experts in their field, better oversight and accountability needs to be put in place within the scientific community.

The reason is that Lynn’s flawed research was passed on for so long by academia without proper scrutiny. This ugly narrative has thus penetrated not only academic discourse but also AI systems.

Rutherford believes the problem is not with AI, but with academia. He goes on to warn that if the right questions are not asked, human and artificial intelligence will leak false information.