GLAAD Report Finds Some AI Chatbots Giving Harmful and Biased Advice to LGBTQ Users

A new report from GLAAD is raising concerns about how some artificial intelligence chatbots respond to queer users. The LGBTQ advocacy organization found that some large language models, or LLMs, reinforced harmful stereotypes, used offensive language to describe a gay person, and in one case, even recommended conversion therapy.
Communication studies professor Samuel Jay of MSU Denver said the findings reflect the data AI systems are trained on. Jay said, ""If you look at the training data that is being used to sort of train these language models, let's not pretend that if it's pulling from the internet, that the internet doesn't have a lot of garbage on it.” Jay also urged businesses to carefully evaluate AI tools before using them for hiring or other major decisions, warning that unchecked bias could expose companies to lawsuits as AI regulation continues to evolve. Colorado lawmakers have already begun addressing AI bias. Democratic State Representative Brianna Titone co-sponsored a 2024 bill requiring AI companies to reduce discrimination. A revised law now gives consumers the right to request a human review of AI-assisted decisions involving jobs, loans, and health care. Titone said AI bias affects more than just the LGBTQ community, citing concerns about racial discrimination as well. She said, "The whole internet is becoming a disaster...It's full of bad information that's hard to discern what is correct and what is not, and people are believing it."

Видео GLAAD Report Finds Some AI Chatbots Giving Harmful and Biased Advice to LGBTQ Users канала Queer News Tonight