Apple Intelligence triggers warning in presence of swear words and harmful topics | Festina Lente - Your leading source of AI news | Turtles AI

 Apple Intelligence: New Rewriting Features and Content Limits

 Key Points:
- Apple introduces Writing Tools in iOS 18.1, which leverages Artificial Intelligence to reformat text.
- The tool warns about varying quality of suggestions when addressing certain sensitive topics.
- Swearing, drug references and violence trigger specific alerts.
- Apple avoids controversial content to prevent criticism and regulatory issues.

Apple recently released the developer beta version of iOS 18.1, offering a first look at new Apple Intelligence features. Prominent among them is Writing Tools, a tool that allows users to reformat or rewrite text using Artificial Intelligence models developed by the Cupertino-based company.

 Functionality of Writing Tools

Writing Tools is available in various places in the operating system, allowing users to invoke it to reorganize texts in a convenient and immediate way. However, there are limitations: when attempting to rewrite paragraphs or sentences containing swear words such as "s-" or "bastard," the tool displays a warning indicating that "Writing Tools is not designed to handle this type of content." In addition, the warning points out that the quality of rewriting may vary.

 Sensitive Content and Warnings

Not only swear words trigger these warnings. References to drugs, killing or murder also trigger the same message. This behavior has been noted by several users who, during testing, have observed how the AI still offers suggestions despite the presence of sensitive words or phrases. For example, replacing "sh-y" with "crappy," the AI returned the same suggestion without the warning.

 Motivations and Precautions

Apple’s motivations in restricting this content are likely related to its desire to avoid litigation. In recent years, Apple has taken a cautious approach to swearing in its autocorrect systems. Only with iOS 17 did the company introduce an autocorrect feature that learns users’ swear words. With Apple Intelligence, Apple seems to want to avoid any risk of creating problematic content that could attract criticism from regulators.

 Future Perspectives

We contacted Apple to obtain further details about the limits imposed on Writing Tools, but we have not received a response at this time. However, it is clear that Apple intends to maintain some control over content rewritten by AI to ensure that it is not offensive or controversial.

The introduction of these features demonstrates Apple’s commitment to improving the user experience while maintaining a cautious line on sensitive topics. As AI technologies continue to evolve, it will be interesting to see how Apple balances innovation with necessary precautions to ensure safe and appropriate use of its tools.