Why trust Greenbot
A free Chrome extension launched by Hiya aims to tackle the growing threat of artificial intelligence (AI) generated deepfakes. With 99% accuracy, Deepfake Voice Detector can identify manipulated audio or video content. This new extension assigns an authenticity score to audio clips, offering a practical defense against misinformation for free.
Deepfake content has become a significant concern as AI generators become more powerful. The technology has been increasingly used for misinformation, scams, and even in attempts to sway political events.
“Deepfake scams can lead employees to share confidential company information or expose critical IT system passwords,” said Kush Parikh, President of Hiya. “The consequences of falling for these scams are immense, especially as vishing is increasingly used with deepfakes to extort or blackmail individuals.”
To combat this, Hiya has made its Deepfake Voice Detector available for free on the Chrome Web Store since October 17. The tool requires only one second of audio to determine authenticity.
Its real-time results and multi-language support make it practical in fast-paced environments. The tool is particularly useful to journalists and newsrooms, as well as businesses that rely on the credibility of online content.
Media and consumer fact-check organizations, including AFP Fact Check and TrueMedia.org, have already trusted Hiya’s extension to validate content before it reaches audiences. Microsoft’s vice chair and president, Brad Smith, praised the tool as an “excellent example of using good AI to combat bad AI,” emphasizing the role of responsible technology in safeguarding information integrity.
The tool’s ease of use and accuracy make it a key asset in the fight against misinformation. While it is free, the daily credit is limited to 20 to prevent overburdening the system. As scams involving deepfakes grow more sophisticated and accessible, the Deepfake Voice Detector provides a much-needed way to stay ahead of AI-driven misinformation.
Recent survey data from Hiya and OnePoll highlights the extent of the issue it’s trying to solve. Among 2,000 respondents in the US and UK, 23% reported experiencing an audio deepfake, while 13% believed they had been exposed to a video deepfake. Of those who encountered video deepfakes, 39% said they found them on social media platforms. In comparison, 15% had experienced them through work-related channels, such as email, video calls, or Slack.
The introduction of this tool marks another step in a broader movement where AI is being used against AI. While deepfake technology continues to evolve, Hiya’s extension allows users to discern what is real and fake, giving some control back to individuals navigating the complexities of digital information. As social media and online platforms become primary news sources, the importance of tools like these will only grow.
