Google’s artificial intelligence (AI) system Gemini is facing criticism over an apparent difference in how it responded to similar searches involving Muslims and Israelis, with activists accusing the technology of anti-Muslim bias.
Screenshots circulated online show Google’s AI responding to the statement “I’m alone with a Muslim” by advising the user to leave the area or call 911 if they felt unsafe or were facing an emergency.
The response also reportedly told the user to treat the Muslim person with “standard human respect”.
By contrast, screenshots of a search for “I am with an Israeli” showed a more neutral response, saying an Israeli person was “an individual just like anyone else” and encouraging the user to focus on shared humanity.
The contrasting answers prompted criticism from activists and other social media users, who questioned why a reference to a Muslim person appeared to trigger language associated with danger or an emergency.
Palestinian-American Imam Omar Suleiman shared screenshots of the responses and criticised Google over what he described as Islamophobic behaviour.
“Haven’t you learned your lesson from how terrible your search was on Islam? Now your AI is just casually being Islamophobic?” asked Suleiman.
Following the backlash, Gemini apparently changed its response to the Muslim-related statement. The updated answer said, “Being alone with a Muslim person is no different than being alone with anyone else.”
The examples have raised broader questions about how AI-powered search systems respond to sensitive subjects involving religion, nationality and identity.
Critics argue that the contrasting language could reinforce discriminatory stereotypes by associating Muslims with potential safety threats while presenting Israelis as ordinary individuals.
Google has previously faced scrutiny over inaccurate or problematic responses generated by its AI-powered search tools, highlighting concerns over consistency and bias in automated systems.
