- The A.I. Ethicist
- Posts
- [CYBERCHECK AI] - The Uncredentialed AI Investigator
[CYBERCHECK AI] - The Uncredentialed AI Investigator
[CYBERCHECK AI] - The Uncredentialed AI Investigator
The Synopsis:
Cybercheck AI, founded by Adam Mosher, is an AI assistant that is used to help law enforcement with investigations. It scours the internet for any information about the suspects, such as social-media platforms to create profiles of them to aid in criminal cases. Additionally, it has been heralded to have an accuracy of more than 90% in retrieving information and creating the profile(s) to aid in investigations. Cybercheck AI has been used in serious cases, according to the article, yet judges and lawyers question its reliability and ethical usage.1
The Analysis:
Adam Mosher claims that Cybercheck AI has more than 90% accuracy in garnering results, yet the methodology of this garnering is opaque; there is not evidence of the large language model, the weights, the biases, and open-source information that Cybercheck’s algorithm is utilizing. He claims that, since the software is proprietary, he does want it to be audited, which I believe is asinine. Adam Mosher has created a software that determines someone’s freedom, and he does not have the decency to audit his own claims and software. I would categorize his software as high-risk, and I believe this software, until vetted and audited by independent agencies, should be removed from use. Although the procedural due process has its imperfections, it is the best of what we have; a notice, an opportunity to be heard, and an impartial tribunal are the minimum in the due process2 , and these facets of the jury process should continue to be upheld. We cannot outsource our better judgment, our minds, and beliefs wholly to AI systems, especially ones that can permanently destroy someone’s life. As I have written in previous blogs, AI can be a tool or weapon; a tool to make our lives more efficient or a weapon to worsen vices, such as racism, sexism, climate insensitivity, and other vile phenomenons. If we need to wait another year or two to mature an AI system before releasing, then that is 100% okay with me; I do not believe we should the rush the process in creating AI systems, lest we use it to destroy ourselves.
The Endnotes:
1 Tim Stelloh, “An AI tool used in thousands of criminal cases is facing legal challenges”, NBC News, accessed May 10, 2024,
2 Nathan S. Chapman and Kenji Yoshino, “The fourteenth amendment due process clause”, National Constitution Center, accessed May 10, 2024,
Reply