Skip to content
Sprint projectAug 26, 2024

Phish Tycoon: phishing using voice cloning

Craig Albuquerque, Melwina Albuquerque · Team Phish Tycoon

Submitted to AI capabilities and risks demo-jam. Sprint projects are early-stage work by participants, not Apart Research publications.

This project is a public service announcement highlighting the risks of voice cloning, an AI technology capable of creating synthetic voices nearly indistinguishable from real ones. The demo involves recording a user's voice during a phone call to generate a clone, which is then used in a simulated phishing call targeting the user's loved one.

Reviews

Judging this Sprint?

Review this project

Your public critique appears on this page without your name. Your private critique is not published; only the Apart team reads it. If you agree below, we share your review with grantmaking.ai (opens in new tab) and the Transformative AI Fund so strong projects can be funded.

Not shown on this page.

Shown on this page, without your name.

Only the Apart team reads this, and funders if you agree below.

Share my name publicly on grantmaking.ai *
Share my private critique with funders *

  1. Nice idea! Showing the entire workflow of audio collection→voice clone→scam phone call is quite helpful. I do recommend using better TTS and voice cloning. As is, the clone doesn’t really sound like you and the robotic TTS decreases the impact of the demo.

  2. I really liked it. I would suggest using a better TTS during the conversational part, as this would have allowed for better immersion and to project what it could look like to have a conversation with someone over the phone only to have your voice stolen.The voice cloning was quite impressive although a bit off. With some polish I feel like this demo could really shine.

  3. Fantastic project! This is a near-term risk, presented in a visceral way! I love the conciseness of the website, and the reveal of the generated audio message. Next steps for this project would be polish - you could use a better voice for the AI conversation (e.g. an ElevenLabs voice), the voice cloning could maybe use additional tuning too, and with more time you could improve the website presentation overall. Great work on this!

Cite this project

@misc{albuquerque2024phish,
  title = {{Phish Tycoon: phishing using voice cloning}},
  author = {Craig Albuquerque and Melwina Albuquerque},
  year = {2024},
  month = aug,
  note = {Submitted to AI capabilities and risks demo-jam, an Apart Research Sprint},
  howpublished = {\url{https://apartresearch.com/sprints/projects/phish-tycoon-phishing-using-voice-cloning}},
  url = {https://apartresearch.com/sprints/projects/phish-tycoon-phishing-using-voice-cloning}
}

Build something like this at the next Sprint

AI Collusion Research Sprint · Oct 23 - 25, 2026