The controversy surrounding a Hong Kong secondary school student’s award-winning AI project, Medisafe, continues to escalate. Two institutions — the Hong Kong Academy for Gifted Education and the Hong Kong New Generation Cultural Association — have publicly denied allegations that the student outsourced the core development of the platform, despite mounting evidence that suggests otherwise.
In a joint statement issued on June 20, the institutions asserted that the student met all eligibility criteria for the Geneva Invention Exhibition and that no outsourcing was involved. They claimed that the U.S.-based AI Health Studio only began providing services after the project had won awards at both the 2024 Hong Kong ICT Awards and the Geneva event, and that the vendor’s role was limited to commercialisation efforts. They also stated that the project data was composed of information from online drug databases and fictional patient profiles, ensuring no breach of patient privacy.
However, these claims appear inconsistent with a now-deleted case study by AI Health Studio — archived by netizens — which clearly outlined the firm’s involvement in the Medisafe project, including user interface and backend development for the student’s father’s medical centre. Screenshots from the submission also show that the app version demonstrated at Geneva already included refinements attributed to the vendor.
The student’s own interview on RTHK further undermines the institutional explanation. In the interview, she claimed to have collaborated with two medical centres and said the system had been tested by over 70 doctors. She also described a technical hurdle as “handling patient databases,” yet struggled to explain what that entailed. Her vague comments about teaching an LLM to “generate prompts to identify mistakes” suggested limited technical understanding, raising further doubts about her authorship.
Hailey, the original whistleblower who faced threats and harassment after exposing the case, has since uncovered additional contradictions. By carefully reviewing the student’s RTHK interview, Hailey highlighted multiple technical inconsistencies and statements that deviate from the project’s claimed development timeline. These findings further question the authenticity of the student’s technical contributions.
Separately, new concerns have emerged around copyright infringement. The Medisafe app lists Rxlist, Drugs.com, and WebMD as sources for its drug database. However, these platforms do not offer public APIs, indicating that data may have been scraped in violation of their Terms of Service. Such actions may also constitute copyright infringement, especially if the information was reused in a commercial or competition context — potentially a serious legal and ethical issue.
The student is also set to present the same Medisafe project at the upcoming Hong Kong Science Fair 2025, to be held at the HKCEC on June 28–29. As of now, there are no signs of a reassessment of the project’s eligibility, despite its involvement in ongoing scrutiny.
Critics argue that the joint institutional statement appears to sidestep key contradictions and serves to deflect reputational damage. As the scandal unfolds, the Medisafe saga is fast becoming a case study in academic accountability, the ethics of AI development, and the responsibilities of institutions overseeing youth innovation competitions.
