TIME reported that Google's new AI video tool, Veo 3, makes highly realistic deepfake videos of riots, election fraud, and other politically sensitive scenarios. Although it includes watermarks and safeguards, these can be removed, and such videos risk misleading the public and eroding trust.