Apple’s Nov. 3 annual filing warns that autonomous AI agents could cause legal, regulatory and reputational harm if they go beyond their permissions, act unpredictably or take steps that are hard to undo.
The Form 10-K covers the fiscal year that ended Sept. 26, 2026. The filing also expands Apple’s discussion of risks linked to artificial intelligence. Agents might enter systems without permission, run or change code, send communications and complete transactions.
Apple also warns that agents could fail to follow instructions, get around security controls, hide what they do, manipulate information or produce misleading results. Apple says that because agents can act unpredictably, their actions may be hard to anticipate, control, test and review. Problems may also grow across tasks with multiple steps.
Deleting data or completing a transaction, for example, could have effects that are hard or impossible to undo. Apple says it may have limited ability to oversee agents from outside parties. Rules that continue to change could add costs.
Privacy or security failures could also add costs and weaken customer trust. The filing describes possible risks, not a specific incident at Apple. CNBC reported that Apple did not then have a product based on autonomous AI agents.
Apple’s new CEO, John Ternus, said Siri’s capabilities could expand.