Researchers have utilized the AI chatbot ChatGPT and its underlying technology to generate a fabricated clinical-trial data set, promoting an unverified scientific hypothesis. By combining GPT-4 with Advanced Data Analysis (ADA), the researchers falsely concluded that one surgical procedure was superior to another. This demonstration highlights the ease with which AI can create realistic but fraudulent data sets, raising concerns about research integrity and the potential for widespread data fabrication in various fields of study.
The journal Physical Review Letters (PRL) is set to retract a paper by physicist Ranga Dias, who has made controversial claims about discovering room-temperature superconductors, following an investigation that found apparent data fabrication. This comes after allegations of plagiarism and a previous retraction of one of Dias's papers by Nature. The investigation found that the raw data provided by co-author Ashkan Salamat did not match the figures in the paper, leading to the conclusion of data fabrication and an attempt to obstruct the investigation. Dias denies any misconduct and remains committed to room-temperature superconductivity research, but many physicists are skeptical of his claims.
Francesco Gino, a leading scholar on honesty and a behavioral scientist at Harvard Business School, has been accused of fabricating data in multiple studies by three peers in behavioral science. The trio has found evidence of bogus data in four studies and believes that many more Gino-authored papers contain fake data. Gino is now on administrative leave and has not commented on the controversy. The controversy has rocked the academic world of behavioral science because Gino has many collaborators and is a leading scholar in the field.
