A flaw in the trust model of a popular artificial intelligence (AI)-powered development environment can allow a malicious actor to modify already approved code extensions for a project and conduct ...