My "Devil Twin" or my discovery of the concept of mirror twin (mirror DNA)
All you have to do is take the raw data file from a DNA test, upload it to Gemini Advanced, for example, and input a prompt asking it to change all nucleotides. All As into G, all Gs into A, all Ts in C, all Cs in T ((this is just one example among countless possibilities). At this stage, the AI performs the modifications and generates a new file with the requested changes in less than thirty seconds. It’s then possible to retrieve this file, create a new account with a DNA testing provider and upload the new file. This will create an artificial genetic profile. As a reminder, my goal with this file was simply to generate a 'mirror twin' of myself, and thus see what it might look like. What I had not anticipated was that this virtual version of myself, in mirror form, actually existed and that this 'mirror twin' had real nephews and half-brothers. One with 91 shared segments for approx. 2000 cM. When I discovered the vulnerability and its potential security and ethical implications, I immediately contacted Google's Bug Bounty program, as well as the companies involved: Ancestry.com, 23andMe, MyHeritage, and MyLivingDNA. It’s now possible to create an UNLIMITED number of versions and profiles from a single DNA data file and upload them to certain sites. This opens the door to numerous frauds! While I am not an expert in computer security, I asked the companies concerned to fix the vulnerability in their files and strengthen their security. To date, no changes have been made.
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2024-10-12
- DOI
- https://doi.org/10.5281/zenodo.15530309
- Primary Topic
- Law, AI, and Intellectual Property
- Type
- article
- Field-Weighted Citation Impact
- 0.00