Jump to content

Undetectable.ai

From Wikipedia, the free encyclopedia
Undetectable AI
Developer(s)Bars Juhasz
Christian Perry
Devan Leos
Initial releaseMay 1, 2023; 15 months ago (2023-05-01)
Written inPython
PlatformCloud computing
Websitewww.undetectable.ai

Undetectable AI (or Undetectable.ai) is an artificial intelligence content detection and modification software designed to identify and alter artificially generated text, such as that produced by large language models.[1][2][3]

History

[edit]

Undetectable AI was developed by Bars Juhasz, a PhD student from Loughborough University, along with Christian Perry[4] and Devan Leos. It was publicly released in May 2023.[5][6]

Reception and analysis

[edit]

Undetectable AI has been discussed in technology and news outlets such as TechTudo[7] and The Inquirer.[8][9]

Academic research

[edit]

Several studies have examined Undetectable AI:

  • In July 2023, researchers from Magna Græcia University tested Undetectable.ai against generative-text and plagiarism detection software. They found that texts processed through Undetectable.ai were significantly harder to detect as AI-generated.[10]
  • In November 2023, Erik Piller of Nicholls State University published a paper examining ethical implications of Undetectable AI in professional writing contexts.[11][12]
  • In August 2023, researchers led by Dr. Christoph Bartneck investigated how Undetectable.ai might affect data quality in online questionnaires. They found that while AI detection systems could identify ChatGPT-generated text, they were less successful with text processed by Undetectable.ai.[13]

Usage and impact

[edit]
  • In November 2023, EarthWeb used Undetectable.ai alongside GPTZero to analyze celebrity apology statements.[14][15]
  • In January 2024, SourceFed announced plans to use Undetectable.ai for AI content detection.[16]
  • A January 2024 report listed Undetectable AI as the 35th most visited AI software in 2023 out of 150 analyzed.[17]

Mechanism

[edit]

Undetectable.ai uses adversarial techniques to modify AI-generated text aiming to make it appear as if a human wrote it.[18]

See also

[edit]
[edit]

References

[edit]
  1. ^ Staff, Knewz (2023-11-01). "New AI Mimics Real Writing — 'No One Can Tell'". Knewz. Retrieved 2024-01-18.
  2. ^ Arasa, Dale (2024-02-06). "How to beat AI detectors". INQUIRER.net. Retrieved 2024-02-21.
  3. ^ "AI took their jobs. Now they get paid to make it sound human". www.bbc.com. Retrieved 2024-06-17.
  4. ^ "How This CEO Makes ChatGPT Speak". SWAGGER Magazine. 2024-02-29. Retrieved 2024-02-29.
  5. ^ Staff, OK! (2023-11-28). "Researcher Working With Royal Air Force Created an 'Undetectable' AI". OK Magazine. Retrieved 2024-01-18.
  6. ^ "Undetectable AI helps emulate 'human' side to AI". KGET 17. May 22, 2023. Retrieved 2024-01-19.
  7. ^ "Undetectable.AI: como saber se um texto foi escrito pelo ChatGPT". TechTudo (in Brazilian Portuguese). 2023-10-10. Retrieved 2024-01-18.
  8. ^ Arasa, Dale (2024-02-06). "How to beat AI detectors". INQUIRER.net. Retrieved 2024-08-08.
  9. ^ Staff, Radar (2023-11-10). "Alan From Mighty Med Condemns AI Cheats — Then Explains How To Cheat With AI". RadarOnline. Retrieved 2024-01-19.
  10. ^ Taloni, Andrea; Scorcia, Vincenzo; Giannaccare, Giuseppe (2023-08-02). "Modern threats in academia: evaluating plagiarism and artificial intelligence detection scores of ChatGPT". Eye. 38 (2): 397–400. doi:10.1038/s41433-023-02678-7. ISSN 1476-5454. PMC 10810838. PMID 37532832. S2CID 260434915.
  11. ^ Piller, Erik (November 4, 2023). "The Ethics of Nondisclosure" (PDF). Rupkatha Journal. ISSN 0975-2935.
  12. ^ Bardard, Neil (2024-01-17). "Promise or Reality? An era of Undetectable AI". Capitol Hill Times. Retrieved 2024-01-18.
  13. ^ Lebrun, Benjamin; Temtsin, Sharon; Vonasch, Andrew; Bartneck, Christoph (2024). "Detecting the corruption of online questionnaires by artificial intelligence". Frontiers in Robotics and AI. 10. doi:10.3389/frobt.2023.1277635. ISSN 2296-9144. PMC 10869497. PMID 38371744.
  14. ^ Buckler, Nicole (2023-11-20). "Top 10 Celeb Apologies Accused of Being Written by AI: Matty Healy, Doja Cat, Joe Rogan, and Elon Musk". The Chainsaw. Retrieved 2024-01-18.
  15. ^ Life, Hollywood (2024-03-20). "Why Celebs Are Using Undetectable AI". Hollywood Life. Retrieved 2024-03-21.
  16. ^ Staff, SourceFed (2024-01-20). "Why SourceFed is Partnering with Undetectable AI". SourceFed - All the News That Matters. Retrieved 2024-02-21.
  17. ^ Rossum, Daan van. "[Report] Generative AI Top 150: The World's Most Used AI Tools (Feb 2024)". FlexOS. Retrieved 2024-04-11.
  18. ^ Grant, H. "Want To Make Your AI Content Undetectable? There's An App For That | Scoop News". www.scoop.co.nz. Retrieved 2024-03-15.