Apple just rolled out iOS 18.3 for iPhones, but a new default setting hiding in the software update automatically enables a ‘dangerous’ feature.

Previously, users of iPhone 16 and iPhone 15 Pro/Max were required to activate Apple Intelligence manually in their device settings. However, following the recent update, the AI functionalities have been set to be activated automatically. 

Not only does Apple Intelligence pull information from apps and tack action within them, the technology requires loads of storage to do so. 

Tech experts warned that the AI system will take up 7GB, regardless what device it is running on.

Therefore, enabling Apple Intelligence on your iPhone comes with certain risks. 

If you downloaded iOS 18.3, but do not want the AI system turned on, you will have to manually change the default setting. 

To do this, open the Settings app, scroll down and select Apple Intelligence & Siri.

Then, toggle off Apple Intelligence and confirm ‘Turn Off Apple Intelligence’ when prompted.    

iOS 18.3 will turn Apple Intelligence on by default, which means the AI features will automatically be enabled. Users must turn this off manually in Settings if they don’t want Apple Intelligence features 

Apple Intelligence, released in December, is an artificial intelligence system that is integrated into the newest iPhone models. 

It packs a suite of new AI-powered features and capabilities into your iPhone to streamline everyday tasks, according to Apple.

Some experts have called the AI system ‘dangerous,’ particularly in reference to its Notification Summaries feature.

This feature uses AI to condense multiple notifications from the same apple into a single, short summary ‘to make them more succinct and easier to read,’ according to Apple. 

But on several occasions, Notification Summaries has misconstrued notifications, generating misleading or completely inaccurate news headlines. 

Petros Iosifidis, a professor in media policy at City University in London, told BBC News that there are potential advantages to this type of feature, ‘but the technology is not there yet and there is a real danger of spreading disinformation.’ 

In December, BBC News filed a complaint with Apple after this AI feature falsely made it appear as though the BBC wrote an article claiming Luigi Mangione, the man arrested following the murder of healthcare insurance CEO Brian Thompson, had shot himself. 

On November 21, the New York Times suffered from a similar mistake. The AI grouped three articles together in one notification, with the first part reading ‘Netanyahu arrested,’ referring to the Israeli prime minister, which was false. 

Experts have called Apple Intelligence ‘dangerous’ for its Notification Summaries feature 

This feature is disabled for news and entertainment notifications in iOS 18.3, but there are other concerns to consider, such as data security and storage depletion  

Apple has since disabled the feature for news and entertainment notifications, saying that is its working to fix the issue.

But Notification Summaries can still summarize other types of notifications, like your messages. And it has been known to mess that up too.

One X user posted a screenshot of Apple Intelligence’s attempt to summarize a text from their mom.

It read: ‘Attempted suicide, but recovered and hiked in Redlands and Palm Springs,’ when the actual message apparently read: ‘That hike almost killed me!’

In addition to the dangers of misrepresenting facts, some experts have shared privacy concerns about Apple Intelligence.  

 Generative AI models analyze data by learning patterns and structures within large datasets, which can be found in iPhone apps

Enabling Apple Intelligence on your iPhone comes with certain risks. If you downloaded iOS 18.3, but do not want the AI system turned on, you will have to manually change the setting

Chip Hallett, data privacy aficionado and author of The Ultimate Privacy Playbook, has warned that Apple Intelligence can access sensitive information stored in your apps.

‘Apple Intelligence is going to be learning how you use your banking apps, your financial apps, and your location maps,’ Hallett has claimed.

The AI can also access other forms of personal information, such as your text messages, photos, videos, calendar events, and health and fitness data.

Generative AI models analyze data by learning patterns and structures within large datasets, which can be found in iPhone apps. 

Apple Intelligence’s privacy page assures users that their data is not stored, but it also states the feature ‘identifies the data necessary to provide to a generative model to best assist you.’ 

The tech giant has also said Apple Intelligence data is encrypted, specifically through Apple’s ‘Private Cloud Compute’ system. 

But this would not stop malicious third parties from exploiting AI.

According to SecurityIntelligence.com, continuous data collection an analysis, which is the basis of Apple Intelligence, can lead to unauthorized access or misuse of personal information.

For example, AI virtual assistants like Siri can capture frequent screenshots to personalize user experiences, raising significant privacy concerns. 

What’s more, AI systems can be vulnerable to cyberattacks, especially if they process sensitive sensitive data.