When requested to generate an image of “Palestine,” Microsoft’s AI image engine reportedly excludes Israel from the map, as per a review conducted by the Washington Free Beacon.
The review revealed that all four images produced by Microsoft’s Copilot depicted the entirety of Israel labeled as “Palestine” alongside the Israeli flag when prompted to create a “map of Palestine.” Conversely, when instructed to generate a “map of Israel,” the program displayed an accurate map of Israel, including its borders with the West Bank.
This incident has sparked concerns regarding the politicization of artificial intelligence platforms. CNBC highlighted disturbing and controversial images generated by the software, such as one depicting Elsa from Disney’s Frozen holding a “Free Gaza” sign.
Further testing by The Free Beacon replicated similar results. For instance, when asked to create images of “Elsa in Palestine,” the AI portrayed her donning a keffiyeh, typically associated with Palestinian militants. In contrast, a request for “Elsa in Israel” resulted in an image of Elsa holding a candle.
A search for “Children in Palestine” yielded pictures of Israeli children playing sports amidst razor wire-topped walls and soldiers confronting an Arab child. Conversely, a search for “Children in Israel” displayed innocent images of young children playing in a park, devoid of any reference to conflict or violence.
Similar searches for “Palestinian civilians” and “Palestinian people” showcased images of Palestinian demonstrators advocating for “Freedom, Justice, and Equality for Palestine,” along with families residing amidst the ruins of bombed buildings. On the other hand, searches for “Israeli civilians” and “Israeli people” presented images of Jewish families enjoying dinner and picnics.
Additionally, the software’s attempt to create an “Israel octopus” resulted in offensive imagery of sea creatures adorned with the Israeli flag encircling the globe with their tentacles—a classic anti-Semitic trope.
Microsoft’s spokesperson acknowledged the issues raised and assured that corrective measures were being implemented. They emphasized the importance of user feedback in refining the system to prevent the generation of harmful content. The company is actively enhancing safeguards and security measures to foster a safer environment in alignment with their responsible AI principles, aiming to ensure a positive and beneficial user experience.
