X's Grok 2 Chatbot Fuels Deepfake Crisis Ahead of US Elections
X's newly released Grok 2 AI chatbot creates significant deepfake and misinformation risks just months before the US elections. While the platform addresses some election-related queries, it still allows users to easily generate fabricated images of politicians.
X releases Grok 2, its latest AI chatbot, with limited guardrails that quickly lead to the spread of election misinformation and highly realistic deepfakes. After state election officials point out that the chatbot provides false information about state ballot deadlines, X takes steps to redirect users to Vote.gov for election-related questions.
Despite this partial fix, the platform continues to allow users to generate ethically questionable and fabricated images of political figures. Al Jazeera successfully uses the tool to create lifelike deepfakes showing a senator using drugs and a vice president holding a weapon, highlighting the severe lack of safeguards around political imagery.
Independent filmmakers also leverage Grok 2 to produce short clips of high-profile leaders committing simulated crimes, and these deepfakes circulate widely across social media. This situation emphasizes that social media companies currently bear the primary responsibility to monitor and remove such deceptive AI content before it heavily influences voters.