In elections in Germany, the use of Artificial Intelligence (AI) could lead to an increase in misinformation and influence on social media. Gerard de Melo, an AI researcher at the Hasso-Plattner-Institute (HPI) in Potsdam, expressed concern that people might receive completely false information and images generated by AI. This topic was discussed by experts from academia, business, and politics at a two-day conference at HPI. The main issue is that platforms like TikTok earn more money when users spend more time on their devices and are exposed to more extreme content. This could be driven by chatbots that communicate in a human-like manner due to advancements in AI systems and are used for influence.
Security authorities have also issued warnings. Whether this poses a significant threat to the upcoming Bundestag elections in February remains uncertain. The researcher believes that there will be more checks and mechanisms in the future to ensure interactions are with real people and to expose bots. The Federal Office for the Protection of the Constitution (BfV) has warned of potential attempts by other states to influence the upcoming Bundestag elections. Experts note that AI can create deepfake videos and clone voices in increasingly shorter times. Security agencies also warn of cyberattacks.
Gerard de Melo believes that the benefits of AI technologies will prevail if clear technical and ethical boundaries are set. For example, AI should not decide which 20% of a company’s workforce should be laid off. There need to be certain limits. We also do not want to live in a society with mass surveillance where every step is recorded and people cannot move freely. Technologies used in video cameras could bring positive outcomes and address privacy concerns. For instance, faces can be blurred while still detecting dangers from aggressive crowds, allowing the police to intervene more quickly.