Meta recently released its Llama 4 AI models, promising a more personalized experience for users. However, the technology has faced criticism for its rightward tilt in design. GLAAD revealed that Meta AI recommended conversion therapy as a possible therapeutic remedy in a series of tests conducted earlier this month.
Meta’s AI Recommendation Sparks Outrage
“Meta AI is legitimizing the dangerous practice of so-called ‘conversion therapy,'” GLAAD posted to social media. The organization’s tests showed that the AI suggested conversion therapy and even recommended several purveyors of the discredited practice. While Meta provided a caveat that the treatment is controversial, critics argue that the AI’s recommendation is not an accident.
Concerns About Bias and Ideology
Meta attributed the rightward tilt to the types of training data available on the internet. “It’s well-known that all leading LLMs have had issues with bias — specifically, they historically have leaned left when it comes to debated political and social topics,” Meta explained. The company claims its goal is to remove bias from its AI models.
Criticisms and Accusations
Some experts see Meta’s claims as an attempt to curry favor with the Trump administration and its right-wing views. Alex Hanna, director of research at the Distributed AI Research Institute, told Axios that it’s “a pretty blatant ideological play to effectively make overtures to the Trump administration.” GLAAD criticized Meta’s attempt to normalize conversion therapy, stating that “both-sidesism that equates anti-LGBTQ junk-science with well-established facts and research is not only misleading – it legitimizes harmful falsehoods.”
Broader Context of Meta’s Challenges
Meta has faced other criticisms recently, including ending its anti-hate speech policy on Facebook and discontinuing fact-checking on the platform. These moves have raised concerns about the company’s commitment to combating misinformation and protecting vulnerable communities.
The controversy surrounding Llama 4 highlights the ongoing challenges tech companies face in developing AI that is both effective and fair. As AI becomes increasingly integrated into various aspects of life, ensuring these systems are designed with careful consideration of their potential impact is crucial.
