OpenAI releases deepfake detector to combat election manipulation

From Daily Mail: 2024-05-12 12:54:59

OpenAI has unveiled a ‘disinformation detector’ to spot deepfakes and combat election manipulation. The tool correctly identifies 98.8% of images created by its DALL-E generator but doesn’t work with other AI programs. This initiative is a response to fears raised by the potential impact of artificial intelligence on elections and society. AI companies are under pressure to take responsibility for the content generated by their technology.

As deepfakes become more convincing and harder to detect, OpenAI’s efforts to develop a deepfake detector tool are crucial. The AI company is collaborating with experts to enhance the technology and address the growing threat of manipulated content. OpenAI aims to lead the charge against disinformation by providing researchers with tools to combat the spread of fake images in real-world scenarios.

OpenAI’s deepfake detector will be a game-changer in the fight against manipulated content. The company’s proactive approach to addressing the challenges posed by deepfakes demonstrates a commitment to safeguarding the integrity of digital content. By joining forces with other industry leaders, OpenAI is taking a step towards ensuring transparency and authenticity in the digital landscape.

To protect the upcoming US presidential election from deepfake interference, tech companies like OpenAI are implementing measures to prevent the creation of fake images of political figures. Midjourney, a popular AI image generator, has blocked users from producing manipulated images of prominent politicians. This move underscores the importance of countering misinformation and preserving the integrity of electoral processes.



Read more at Daily Mail: OpenAI CEO launches ‘disinformation detector’ over fears that faked images will sway presidential election this year