Apple has delayed approval of an email-app update with AI-powered language tools (#ChatGPT), concerned with insufficient measures to prevent the AI from generating content inappropriate for children. The app can raise the age restriction to 17 and older or include content filtering.
From a #technologygovernance perspective, and we're simplifying here - Apple is not taking responsibility for the AI generating inappropriate content for that application class. In the example case, OpenAI (presently) does not own the risk for everyone. Hence the onus is on the app provider to mitigate the #risk of inappropriate content for children. The options are to run your own generative #AI / large language model (#LLM) / ~ChatGPT~ (something that is becoming easier) or employ another robot to watch over the generated content (that is, outsource the risk, not that there's a rich market of such providers yet). Calculating a path for success today and tomorrow is non-trivial.
