Khaberni - Google withdrew the artificial intelligence model "Gemma" from its experimental platform AI Studio, days after US Senator Marsha Blackburn accused the model of fabricating false sexual assault accusations against her.
The company stated in a post on the X platform: "We have noticed reports of non-developers attempting to use Gemma for posing real questions... It was never designed as a consumer tool or for use in this manner, and to avoid this confusion, we decided to discontinue its availability through AI Studio."
Although Google did not directly mention Senator Blackburn by name, the decision followed a wave of harsh criticism that came after an official letter she sent to CEO Sundar Pichai, in which she stated that the model falsely claimed a law enforcement officer accused her in 1987 of coercing him to procure drugs for her and that the relationship involved "non-consensual acts" during her campaign.
Blackburn confirmed that these claims were "completely untrue," noting that her Senate campaign occurred in 1998, not 1987, and that the links listed by the model as sources were either error pages or news unrelated to the topic, adding that these accusations represent "slander issued by a Google-owned model."
In its response to the controversy, Google acknowledged that hallucination in artificial intelligence, where models fabricate incorrect information, remains a significant challenge even in small models like Gemma, which was developed to be a more efficient and sustainable alternative to larger models like Gemini 2.5 Pro.
Gemma is a family of lightweight, open-source models that developers can integrate into their applications, and was available through AI Studio before it was withdrawn.
Despite the stop, researchers and developers can still access it through Google's API (Application Programming Interface).
This incident occurs amidst escalating political debate in the United States over bias in artificial intelligence systems, where several Republican legislators have accused such models of adopting "liberal" orientations. Earlier this year, US President Donald Trump signed an executive order to ban what he called "conscious AI" or "woke AI".
