A Microsoft engineer has raised an alarm over the company’s AI image generator, Copilot, which is said to produce violent and sexual images, and also ignore copyright laws, according to reports. The engineer, Edward Tian, an MIT graduate who worked on the OpenAI project, told officials in the US and the company’s board about his concerns. Tian, who left Microsoft last month, said in a tweet that the company had done nothing to fix the issues he raised, despite having six months’ notice. The concerns come in the wake of the accidental creation of a generation of deepfakes, and the wider issue of the potential misuse of AI technologies.