I came across a pretty interesting post on using APIs for research, and the benefits, and challenges that researchers face when depending on APIs. It was another side of API stability and availability that I hadn't considered too much lately. Social media platforms like Twitter and Facebook are rich with findings to be studied across almost any discipline. I regularly find social media API studies at universities from areas like healthcare and Zika virus, algorithmic intellectual property protection, all the way up to US Navy surveillance programs that are studying Twitter.
APIs are being used for research, but there are rarely API platform plans crafted with research in mind. Flexible rate limits, custom terms of service, that give them access to the data they need. I'm assuming that some companies have behind the scenes deals with some universities, or larger enterprise research groups (IBM, etc), as well as government agencies, and police agencies. The problem with this, is that 1) there is no virtual public front door to walk through and understand research levels of access, and 2) the details of partnerships are not publicly, equitable, and auditable by journalists, and other groups.
The author of this essay provides a lot of details regarding what it is like to depend on APIs for your research. Some of them could put your career in jeopardy if the terms of service, and access levels change before you could finish your research, or dissertation. I'm not sure what the responsibility of API providers should be when it comes to making their resources available for research, but it is something I will be exploring further. I will be reaching out to researchers about their API usage, but will also be helping encourage API providers to share their side of things, and maybe eventually formalize how API providers make their valuable resources available for important research.