I am also opposed to this kind of petition on the grounds that it inhibits standardization of file formats. I work at a startup and we work with medical records

I am also opposed to this kind of petition on the grounds that it inhibits standardization of file formats. I work at a startup and we work with medical records from a number of different hospitals. So much code is dedicated to just dealing with the spaghetti. A major roadblock for our startup is getting medical information integrated into our system. It's difficult to compete with the IBMs and Epics of the world when we don't have a million developer hours to dedicate to writing plugins for every vendor. It's not just us struggling with crappy data management - it confuses hospital staff, too. Our customers are frustrated when we tell them things like "you gave us <XYZ obscure file type> which doesn't contain the information we need; do you have <ABC obscure file type> instead?". MRI scans (DICOMs) are particularly gnarly. Even the IBMs and Epics of the world struggle to not make crappy software. How do you present a relevant medical record to a doctor at the exact right time they need it? There is so much data to sift through that medical information frequently slips through the cracks when patients transfer hospitals or their hospital merges with another. If there isn't a standard way to query an electronic health record then companies are incentivized to just throw data at an LLM to parse it (which is exactly what we're moving towards). Trying to build AI-type solutions will just make these companies even more data-hungry and result in a less reliable solution. I'm not opposed to continuing to hold startups/big tech accountable for keeping personal health information private and secure.