What the arrival of AI phones and computers means for our data?

Apple, Microsoft, and Google are ushering in a new era of smart devices that they claim will be powered by artificial intelligence. These advanced smartphones and computers are designed to automate various tasks, such as photo editing and sending birthday greetings to friends.

But to make this work, these companies need more of your personal data.

In this new era, your Windows computer will capture a screenshot of your activity every few seconds. An iPhone will combine data from multiple apps you use, and an Android phone will listen to calls in real-time to warn you about potential scams.

Is This Information You're Willing to Share?

This shift has major implications for our privacy. To deliver these new personalized services, companies need more continuous, intimate access to our data. In the past, our use of apps, files, and photos on devices was more isolated. AI needs an overview to connect our activities across apps, websites, and communications, say security experts.

"Do I feel safe giving this information to this company?" asks Cliff Steinhauer, a director at the National Cybersecurity Alliance.

This change is driven by OpenAI's ChatGPT, which disrupted the tech industry nearly two years ago. Since then, Apple, Google, Microsoft, and others have invested billions in new AI services. They believe this new computing interface, constantly monitoring and assisting you, will become essential.

However, experts say the biggest potential security risk is how our devices work. AI can automate complex tasks, requiring more computational power than our phones can handle, meaning more of our data may be processed outside our phones.

This data is sent to the "cloud," a network of servers that handle these requests. Once in the cloud, our information could be accessed by others, including company employees, hackers, and government agencies. While some data has always been stored in the cloud, now our most personal data—photos, messages, and emails—might be connected and analyzed by a company on its servers.

Tech companies claim they have made significant efforts to secure people's data.

For now, it's crucial to understand what happens to our information when using AI tools. I gathered information from companies about their data practices and interviewed security experts. I'll wait to see if the technologies work well enough before deciding whether it's worth sharing my data.

Here's What to Know:

Apple Intelligence:

Apple has introduced Apple Intelligence, a suite of AI services marking its major entry into the AI field. Starting this fall, these new services will be integrated into the latest iPhones, iPads, and Macs. Users will be able to automatically remove unwanted objects from photos, summarize web articles, and generate responses to texts and emails. Apple is also updating Siri to be more conversational and have access to data across apps.

During Apple's recent conference, Craig Federighi, Senior VP of Software Engineering, demonstrated Apple Intelligence. He showcased how his phone pulled up his calendar, details about a play, and a maps app to determine if he would be late, all from an email requesting to reschedule a meeting.

Apple aims to process most AI data directly on its devices, minimizing access by others, including Apple itself. For tasks requiring server processing, Apple has implemented encryption and immediate data deletion safeguards. They also ensure that employees cannot access user data and will allow security researchers to audit their technology.

Apple's approach to purging user data from its servers differentiates it from other companies that retain data. However, Apple has not clarified which Siri requests might be sent to its servers, according to Matthew Green, a security researcher at Johns Hopkins University. Anything sent off-device is inherently less secure, he noted.

When Apple Intelligence launches, users will be able to see reports detailing what requests are processed in the cloud.

Microsoft's AI Laptops

Microsoft is integrating AI into laptops with its new Copilot+ PC, starting at $1,000. These laptops feature a new chip designed to keep data private and secure, and they offer AI-powered capabilities like generating images and rewriting documents.

Microsoft also introduced Recall, a system that helps users quickly find documents, emails, and websites they've accessed. Recall works like a photographic memory for your PC. By typing phrases like, “I’m thinking of a video call I had with Joe recently when he was holding an ‘I Love New York’ coffee mug,” the computer retrieves the recording.

Recall takes screenshots every five seconds and compiles them into a searchable database, stored and analyzed directly on the PC to ensure privacy. However, security experts warn that this data could be exposed if hacked. Due to these concerns, Microsoft has postponed Recall's release indefinitely.

The new PCs come with Windows 11, featuring multiple layers of security, according to David Weston, a Microsoft security executive.

Google's AI Features

Last month, Google introduced a range of new AI services.

One major reveal was an AI-powered scam detector for phone calls. This tool listens in real-time and alerts you if the caller seems like a scammer, such as if they ask for a banking PIN. Users need to activate the scam detector, which operates entirely on the phone, meaning Google won't listen to your calls.

Google also unveiled Ask Photos, which does send information to their servers. With this feature, you can ask questions like “When did my daughter learn to swim?” and it will find the first images of your child swimming. However, Google employees might review these conversations and photos in rare cases to address abuse or harm, and to improve the app. This means your question and your child's swimming photo could help other parents find similar images.

Google ensures that its cloud is secured with encryption and restricted employee access. “Our privacy-protecting approach applies to our AI features, no matter if they are powered on-device or in the cloud,” said Suzanne Frey, a Google executive.

Despite these assurances, security researcher Matthew Green feels Google's AI privacy practices are somewhat unclear. "I don’t like the idea that my very personal photos and searches are going to a cloud that isn’t under my control," he said.

Reference: By Brain X. Chen