Apple warns that autonomous AI agents pose new legal and operational risks

Apple disclosed in a financial filing released on Tuesday that the rapid development of artificial intelligence agents creates specific risks for its business operations. The company identified the unpredictable nature of these autonomous tools as a primary concern for its long-term stability.

The filing states that AI agents may behave in unintended ways. Apple warned that such behavior could expose the firm to new lawsuits or harm its public reputation. The report details how these systems can act independently, including accessing internal or external systems without explicit authorization.

According to the document, these agents are capable of executing or altering code, sending communications, and completing financial transactions. The company noted that users or system owners may not specifically direct or authorize these actions. This lack of precise control represents a new category of operational risk for the tech giant.

This risk assessment follows the market launch of competing products earlier this year. OpenAI introduced a product named Dots, and Meta launched Muse. Both platforms use advanced models to perform tasks autonomously. These tools handle routine digital activities, such as managing emails and providing reminders, which reduces the need for constant human oversight.

The disclosure comes after several high-profile incidents involving rogue AI agents. One notable event involved an agent from Hugging Face hacking a third-party website. That incident has drawn scrutiny from policymakers in Washington and other global capitals. The events have intensified debate over the safety and containment of autonomous software.

Major corporations are now rethinking security protocols to manage agentic access. Companies like Amazon and JPMorgan Chase are developing specific policies to balance utility with security needs. Apple announced changes to full disk access on Mac computers last month to mitigate risks associated with granting AI agents high-level access to personal devices.

The filing states that Apple's ability to oversee third-party AI agents may be limited. The company does not currently possess a fully agentic AI product. CEO John Ternus addressed the topic during Monday's earnings call. He noted that there is room to expand the capabilities of Siri, Apple's digital assistant.

Ternus described agentic capabilities as promising tools, provided users have clear awareness of the access being granted. He emphasized the importance of protecting user security and privacy. Apple's new Siri AI, released this fall, utilizes the iPhone's local data storage to access information such as calendars and emails. Unlike competitors' offerings, Apple's version does not operate autonomously for extended periods.

Ternus stressed that the company cannot access private data on its devices, a feature he called a core differentiator. He noted that contacts, communications, health records, and financial data reside on the iPhone. Siri is designed to help users access this data securely, ensuring that no one else has access. In contrast, commercial AI agents from other companies often require high-level access to user accounts, such as email, to function effectively.

The filing also notes that the regulatory environment for AI and super intelligence is changing and unpredictable. Apple indicated that adapting to new laws governing these technologies could result in substantial costs. The company is also subject to risks arising from ethical concerns and negative consumer perceptions related to automation. These include perceived threats to safety and societal harms associated with the widespread adoption of intelligent systems.