Science & Tech · AI · 2 hrs ago
Apple warns that AI systems could act without authorization
Apple warned US regulators that AI systems may act without permission in its annual filing.
The company said AI is increasingly part of its products, services and operations.
Those systems could exceed their permissions, ignore instructions or produce misleading results, Apple said.
Because AI can be hard to predict and control, mistakes could lead to actions such as deleting data or making transactions that are difficult to reverse.
Apple said these failures could bring financial, legal and reputational harm and that its ability to oversee its own and third-party AI may be limited.
The company also said laws and regulations for AI are uncertain, and it may face significant costs as rules develop and regulators examine the technology.
Apple warned that autonomous AI agents could take actions without authorization and pose risks to its business.
The company said such systems could expose it to lawsuits, regulatory measures, reputational damage and financial harm.
Apple cautioned that it may be difficult to predict, control, test or audit how AI systems behave.
It also said its ability to oversee and control third-party AI agents may be limited.
Apple said users should understand what access they grant to AI tools and that security and privacy should be protected.
- Who
- Apple, the technology company.
- What
- Apple warned that AI systems, especially autonomous agents, may act without authorization and cause harm to its business.
- When
- Apple disclosed the risks in its annual filing on Tuesday, November 3, 2026.
- Where
- In an annual filing to the US Securities and Exchange Commission.
- Why
- Apple said AI systems may behave unpredictably, while oversight can be limited and laws and regulations are changing.
This story does not have two clearly opposing sides.
OpenAI released Dots, and Meta introduced Muse, AI agents designed to work autonomously on tasks.
Apple announced changes to full disk access on Macs to reduce risks related to AI agents installed on computers.
Apple CEO John Ternus said AI agent capabilities were promising tools and that Apple saw room to expand Siri.
Apple warned in its annual filing that AI agents could behave unpredictably and take unauthorized actions.
- Filing
- Apple’s annual filing to the US Securities and Exchange Commission
- Third-party oversight
- Apple said its ability to oversee and control third-party AI agents may be limited.
- Potential harms
- Lawsuits, regulatory measures, reputational damage and financial harm
- AI agents
- OpenAI’s Dots and Meta’s Muse
- Siri
- Apple’s Siri could access data stored on the iPhone but could not work autonomously for long stretches like Dots or Muse
3 sources
What people are saying
In particular, artificial intelligence agents and other technologies can take actions autonomously, including interacting with internal or external systems that in some cases they may not be authorized to access, executing or altering code, sending communications, and completing transactions.
may behave in unintended or unpredictable ways, including by taking actions that were not directed, instructed or authorized.
Dado que muchas de estas tecnologías son no deterministas, su comportamiento puede ser difícil de prever, controlar, probar y auditar, y los errores pueden acumularse
We just want to make sure that users are aware of what they are giving access to. We want to make sure this is done in a way that protects security and privacy.