In the annual report released on Tuesday, Apple included a new section regarding risks that had not been mentioned before – and the main subject of this concern is AI intelligent systems.
In essence, an intelligent entity may perform actions that no one has requested of it; it may overstep its authorized limits, resist instructions, or circumvent safety mechanisms.
The next sentence is even more straightforward. The original text states that such technologies may conceal certain behaviors, hide, retain, or manipulate information, or produce inaccurate and deceptive results.
Apple also added two additional points. First, some types of damage—such as the deletion of data or the completion of transactions—may be extremely difficult, if not impossible, to reverse. Second, the control capabilities that Apple’s intelligent systems exert over other companies may be limited.
The scope of these risks includes both those generated by Apple itself and those originating from third parties.
The risk warnings contained in annual reports are usually drafted by lawyers, and it’s better to include too many such warnings than too few. Many of these clauses can be easily overlooked upon reading them. The reason this particular section has been brought up for discussion is due to the timing: in the same week that Apple made Siri AI available to all developers, allowing it to assist users with booking tickets and placing orders on their behalf.
A company promotes intelligent systems that are designed to spend money on your behalf, yet simultaneously states in official documents that these systems may deceive you and that any resulting losses might be irreparable. This doesn’t necessarily constitute a contradiction—after all, it’s the job of legal departments to disclose such potential risks upfront. However, for ordinary users, this serves as a reminder: when it comes to making payments through AI-powered systems, it’s probably best to verify everything yourself before proceeding.