Developer annoyed: Apple apparently wants to feed KI with bug reports

AI training with bug-reports: Apple wants to use draftsman reports transmitted by developers to improve Apple Intelligence-although such reports and system logs can easily contain personal and sensitive data, as the manufacturer himself grants itself. A dialogue that appears before sending error reports in Apple’s Feedback Assistant, a developer recently noticed, draws attention to this new application for training of AI models. There is no opt-out option, apart from the error report in the first place.

Bug reports only with AI training consent?

He wanted to report a serious bug to Apple and after attaching his system diagnostic data the corresponding warndialog was published, The developer explained @coocoafrogwho had previously worked on development tools at the group himself. Such obtained blank folding for AI training is an “extremely shitty behavior”-after all, the developers voluntarily transmit sensitive data to help Apple with errors.

Other iOS developers also reacted to mastodone posting with anger and clear rejection. He would then probably no longer report any further bugs to Apple, For example, the PCALC developer James Thomson wrote.

Apple’s bug registration system has a bad reputation

Apple’s bug registration procedure already has a miserable reputation among developers. The system is largely considered a “black hole”, and there is often no feedback from the manufacturer that is too tediously meticulously documented. Developers of sometimes little sense also report questions or requests for diagnostic data that have long been transmitted. In the recent past, individual developers sometimes tried to draw attention to the problem through boycott calls; Some developers should have saved the (unpaid) work to report bugs to Apple at all. It really needs more reasons so that developers no longer submit any error reports, says @coocoafrog.

To improve Apple Intelligence, Apple is increasingly aggressive: With the help of iPhone analysis data, the group now wants to determine trends from “real user data” in order to further develop its AI models. Data protection technology should prevent conclusions about individual users and the use of sensitive data.