New York City Mayor Eric Adams has attracted attention for using artificial intelligence (AI) to generate robocalls in multiple languages, despite not being able to speak those languages himself. Adams revealed this information on Monday, stating that the calls, which promote city hiring events, have been made in languages such as Mandarin and Yiddish. However, there has been no disclosure that the calls were AI-generated or that Adams does not actually speak these languages.
Adams, a Democrat, claims that the robocalls are aimed at reaching the diverse population of New Yorkers. However, this has raised ethical concerns about the government’s use of AI, particularly with regards to deepfakes – manipulated media that make it appear that someone is doing or saying something they are not. The Surveillance Technology Oversight Project, a watchdog group in New York, has criticized the robocalls as unethical and misleading to residents of the city. Albert Fox Cahn, the organization’s executive director, described the mayor’s actions as “deep fakes” and an “Orwellian” use of AI. While acknowledging the need for announcements in multiple languages, Cahn argued that the deepfakes are an unnecessary and eerie project.
The use of AI and deepfakes, especially in politics and election misinformation, has prompted calls for greater regulation from both the government and major media companies. Google has announced plans to apply new labels to deceptive AI-generated political advertisements, while Facebook and Instagram parent company Meta has policies restricting manipulated audio and imagery used for misinformation. In the US Senate, a bipartisan bill has been proposed to ban “materially deceptive” deepfakes related to federal candidates.
In recent weeks, various technology companies have showcased AI tools that can dub a person’s speech in another language, making it appear as if they are speaking fluently in that language. Examples include Spotify, which introduced an AI feature to translate podcasts into multiple languages using the podcaster’s voice, and startup ElevenLabs, which developed a voice translation tool to convert spoken content into different languages while maintaining the original speaker’s voice.
The mayor’s office confirmed that they used ElevenLabs’ tool for the robocalls and ensured the accuracy of translations by having native speakers review the recordings. Calls were made in languages such as Spanish, Yiddish, Mandarin, Cantonese, and Haitian Creole. The technology has also been used to promote concerts organized by the Adams administration. Adams defended his use of AI, emphasizing the importance of being able to communicate with New Yorkers in the languages they understand.
As the government’s use of AI continues to evolve, it is crucial to address the ethical implications and consider regulations to ensure transparency and accountability.
