How to Protect Your Data on All Your Devices

Every year we become increasingly reliant on and connected by technology. Social media helps us keep in touch; Alexa and ChatGPT answer all our questions; digital maps point us in the right direction; and apps make it easy to do…just about everything.

But all that connection can come at the expense of privacy and security. Anyone who uses tech—on any platform—can be affected by a data breach. And, unfortunately, if you’re of a certain age, you're especially likely to be a target of a scam: In 2024 alone, middle-aged and older adults accounted for more than 60% of victims of Internet crime and lost over $9.5 billion to hackers and scammers, per the latest report from the FBI Internet Crime Complaint Center. Hackers are always finding innovative ways to sneak into our lives and steal valuable data that makes up our digital footprint, from posing as tech support people to inundating us with dangerous links.

Luckily, you don’t have to be a tech pro (or anything close) to protect yourself and your loved ones from digital disasters. Our guide can help you learn which threats should be top of mind for you and how you can build a stronger cybersecurity shield in seconds.

✔️ Keep your operating system and apps up to date. When a pop-up says it’s time for another security update on your device, do it! And make sure you have the most recent anti-virus software installed.

✔️ Use multifactor authentication. This creates an extra step when logging onto email, banking, and social media accounts so you prove it’s really you (keeping hackers out) with a code sent via email, text, or an authentication app.

• iPhone/Apple: Go to Settings > Your Name > Password & Security > Turn on Two-Factor Authentication.
• Android/Google: Go to safety.google/authentication​

✔️Create stronger passwords. A password manager likeProton Pass can help. You need to remember only one password; the app generates passwords for all your accounts that are locked inside it. Your next best move is to strengthen passwords by using 12 or more characters and random words and numbers.

Reduce the flow of personal information from your devices with these general steps.

ON YOUR PHONE

✔️ Limit location tracking.

Tell apps to stop collecting digital breadcrumbs on your every move— or to do so only when you actually need GPS.

• iPhone: Settings > Privacy & Security > Location Services > Turn Off, or adjust by app
• Android: Settings > Location > App Location Permissions > Turn Off, or adjust by app
  

✔️ Tell apps to stop asking about tracking.

Give an automatic “no” to tracking requests.

• iPhone: Settings > Privacy & Security > Tracking > Turn Off Allow Apps to Request to Track
• Android: Settings > Privacy > More Privacy Settings > Ads > Delete Advertising ID
  

✔️ Adopt digital minimalism.

Delete apps you don’t use, and download only new ones that score high on privacy via these tools:

• Apple App Store: Review apps’ privacy pledges with Privacy Nutrition Labels.
• Google Play: Check out apps’ Data Safety sections under their descriptions.

ON YOUR VIRTUAL ASSISTANT

✔️ Delete your voice history.

Apple Siri and Google Assistant keep audio recordings only if you opt in to improve their services, but you have to opt out to stop Alexa from keeping them.

• Apple Siri: In Home app, go to Settings > Siri > Siri & Dictation History > Delete Siri & Dictation History
• Google Assistant: Go to your Google Account’s Assistant activity page > Sign into your account > tap More > Delete activity by all time or say, “Hey Google, delete my last conversation/today’s activity/this week’s activity.”
• Amazon Alexa: In-app, tap More > Alexa Privacy > Review Voice History > Select date range, device, or profile to delete or delete all.

ON YOUR COMPUTER AND LAPTOP

✔️ Use a pro-privacy browser or a tracker blocker. Switch to a browser like Safari or Firefox to better protect your privacy, or add a tracker blocker such as Privacy Badger or uBlock Origin to keep more of your data to yourself.

ON AI CHATBOTS

✔️ Delete your chat history.

Many AI chatbots allow you to delete chat history and then schedule chats to be permanently deleted from their backend within 30 days (unless retained for company’s interests, such as security or legal reasons).

• ChatGPT: Settings > Data Controls > Delete All Chats
• Claude: Click Chats > Scroll down to show more > Click checkbox > Trash icon > Delete
• Copilot: Settings & Privacy > Privacy > Data Options > Copilot Activity History > Delete History
• Gemini (Google Deepmind): Menu > Recent > Delete

✔️ Turn off personalization and targeted ads.

To prevent chatbots from remembering details from previous conversations and using them to “personalize” your interactions or contribute to targeted ads, turn this capability off.

• ChatGPT: Settings > Ads Controls > Delete Ads Data
• Copilot: Privacy > Personalization > Turn Off Personalization. For specific memories, ask Copilot: “What do you know about me?” or demand deletion of specific memories: “Forget _.” To delete all memory, select Memory > Delete all memory
• Gemini: In Google Account > My Ad Center > Manage Privacy > Toggle Off Activity On Sites and Apps

✔️ Turn off memory.

To prevent chatbots from retaining specific memories about you or your chat history, turn off memory. (As of reporting, Claude’s memory is automatically off. Double-check your settings using the steps below.)

• ChatGPT: Settings > Personalization > Memory > Toggle off Reference Saved Memories and Reference Chat History and delete specific memories manually
• Claude: Settings > Capability > Memory > Toggle off Search and Reference Chats and Generate Memory from Chat History
• Copilot: Settings > Personalization > Saved Memories > Toggle Off
• Gemini: Settings > Personal Context > Your Past Chats With Gemini > Toggle Off

✔️ Opt out of AI model training.

Many chatbots require that you manually opt out of your data being used to train AI models.

• ChatGPT: Settings > Data Controls > Toggle Off Improve the Model for Everyone
• Claude: Settings > Privacy > Toggle Off Help Improve Claude
• Copilot: Privacy > Training On Conversation Activity > Training On Voice Conversations
• Gemini: In Gemini Web App > Gemini Apps Activity > Turn Off and Delete Activity

✔️ Use incognito mode.

This means your chats won’t be used as training data or retain memories if you’ve otherwise personalized your experience.

However, even in these modes, chats can be stored for anywhere from 72 hours to 30 days and beyond (if, for instance, content is flagged). Ultimately, you know by now: No chat is completely private.

• ChatGPT: Click the speech bubble icon to start a temporary chat.
• Claude: Click the ghost icon to switch into incognito mode.
• Gemini: Click the clock icon to start a temporary chat.

TECH-SPEAK TRANSLATED

Your “data”—the information available online and in apps that includes your age, health status, income, interests, and more—is quickly becoming one of the most valuable assets for advertisers.

Here’s how it works: Much of the most telling data—like where you go with your phone—comes from mobile apps. With these apps, third-party companies put together as much data on you as they possibly can: your phone’s advertising ID (a code used to identify your unique device), your GPS location, and any demographic or personal info you’ve shared with the app and those it’s linked with. They use these data points to assemble a digital dossier and then send that to companies hired by advertisers to target you.

In a fraction of a second, you go up for auction—and the highest bidder wins the opportunity for its ad to appear on your screen, explains Jane Hoffman, a Harvard fellow and author of Your Data, Their Billions: Unraveling and Simplifying Big Tech.

Maybe targeted ads don’t seem like a big deal. “But it matters, because your data can be used against you,” says Hoffman. Data brokers aren’t just in the advertising business; they also sell the power to mine your personal history, to find and follow you, and to predict and influence behavior on a massive scale.

According to the latest report from Duke University’s Cyber Policy program, data brokerage activity in the U.S. is a “virtually unregulated practice.” Justin Sherman, a senior fellow and research lead for the Data Brokerage Project at Duke, told the House Energy and Commerce Committee at a 2023 hearing that the data brokerage system “profits from unregulated surveillance of every American, particularly the most vulnerable.”

Profile collection could be used to discriminate against people, while sensitive information like where you live and work—which is readily available on people-search websites—can be used by stalkers and abusers. And inaccurate data, which is hard to correct, has damaged people’s credit, preventing them from buying houses and getting jobs.

Eight in 10 people are concerned about how much personal data companies are collecting and believe the risks outweigh the benefits, per the latest Pew Research Center study. We deserve default settings that automatically put us in control of our data and allow us to consent meaningfully to how it’s being collected and used. But for now, it’s on us to adjust our settings and choose apps, products, AI tools, and browsers that block digital spies.

When you get the letter or email from a company letting you know you’ve been a victim of a data breach, you might wonder, Should I be concerned about hackers racking up thousands on my credit card or taking out fraudulent loans? And if so, how can I stop them?

After a data breach, your info can be shared and sold on the dark Web or used for identity theft. You can find out how many breaches you’ve been involved in by searching for your email address on haveibeenpwned.com (“pwned,” pronounced “poned,” is gamer-speak for being “owned,” or controlled or coerced into doing something).

Reduce your risk by changing the passwords for compromised accounts, setting up multifactor authentication, monitoring your bank and credit card accounts closely and upgrading your digital security.

Artificial intelligence, or AI, has been in the news recently because of the rapid rise of chatbots like OpenAI’s ChatGPT, Anthropic’s Claude, Google’s Gemini (Deepmind) and Microsoft’s Copilot. It can also be found in self-driving cars, smart home devices and patient portals used by the health care industry.

While these systems bring a host of benefits, there are, of course, concerns. And because AI is becoming more prominent and available, more individuals could be vulnerable in multiple ways.

Unfortunately, it’s “basically impossible” to escape digital surveillance, and AI “may make matters worse,” says Jennifer King, privacy and data policy fellow at Stanford University Institute for Human-Centered Artificial Intelligence, in an interview about her latest report on privacy in the AI era. “AI systems are so data-hungry and intransparent,” she states, “that we have even less control over what information about us is collected, what it is used for, and how we might correct or remove such personal information.”

Essentially, AI supercharges uncertainty and risk around digital privacy and security. Jeff Crume, an engineer and cybersecurity expert, tells IBM that because AI collects so much data, these companies have a “big bullseye” hackers want to hit. AI technology can also be used to automate cyberattacks (between December 2025 and January 2026, Mexico’s government was targeted), steal or accidentally leak sensitive information in data breaches (which happened in early 2026 at Meta), and abuse individuals by mass-producing sexualized deepfake images and videos, to name a few problems (Grok, X’s AI tool, was under investigation in the EU at the time of reporting for producing thousands of image-based sexual abuse images).

For the average individual, it’s hard to know how secure AI is because companies are not transparent about their security protocols or how data is stored and used. Some have even collected data from children—and failed to safeguard it.

Some companies allow humans to review chats, and some allow you to opt out of sharing data to train AI models, but you typically have to opt out manually. Even then, legalese in privacy policies makes it hard to decode just how effective that choice is. Case in point: While data may be “de-identified” (tech-speak for unpaired from a unique individual user’s ID), it can be re-identified if it’s in the company’s interest for legal or other reasons.

Like other digital privacy and security issues, two things are true: 1) There’s much that’s out of the individual consumer’s control and must be addressed with regulations at a higher level, and 2) knowing about certain risks could still put you in a better position to protect yourself. For instance, know that AI-generated information is not always factually correct because models can “hallucinate” or invent facts and figures to sound authoritative. “Dr. AI” may not be any better than “Dr. Google,” and chatbots can give bad or even dangerous medical advice.

It’s also important for your security to understand that talking to AI isn’t like talking to an imaginary friend: It’s designed to keep you engaged and influence you, and words you plug into it don’t poof into nothing; they’re generally stored. The companies themselves, such as Microsoft, explicitly say that you “shouldn’t provide any confidential or sensitive personal data” you would not want them to use, including anything that could reveal “your race, religion, sexual orientation, or health status.”

When it comes to avoiding AI-generated scams, pay attention to the usual social engineering tools used to trick you: Question the sense of urgency, listen to your gut, don’t befriend charismatic strangers on the Internet, and find ways to double or triple-check that the information you’re getting is legitimate.

Also, keep in mind that the information you give AI could sometimes be used to target you with ads in other places—and sponsored content is invading the AI chat. While Anthropic (Claude) is currently ad-free, Meta uses information plugged into their generative AI products to give users targeted ads on Facebook and Instagram. Google AI search results include sponsored products (and Gemini will, too, in the near future), and OpenAI is testing ads on the free version of ChatGPT. The only escape in sight: a paid subscription.

It’s not solely companies we that should be concerned about: The Pentagon and Anthropic are also at war over how AI technology should be used by the U.S. military. At the center of this dispute: concerns about “domestic mass surveillance and mass propaganda" and “fully autonomous weapons,” according to Anthropic founder Dario Amodei’s essay “The Adolescence of Technology.”

It’s all happening so fast it’s easy to get whiplash. As with any new tech, GH experts recommend being cognizant of where you input personal data and staying skeptical about AI-generated advice from chatbots. Sadly, it’s also important to question photos and videos that you see online—especially those that seem too shocking or controversial to be true. Look for legitimate reporting from trusted news outlets to learn whether you’re being exposed to misinformation or the truth.

If you’ve ever talked about something only to see an ad for it pop up on your screen a moment later, you’ve probably at least wondered whether companies could be spying on you. But is this actually possible?

David Choffnes, Ph.D., a computer science professor and executive director of the Cyber-security and Privacy Institute of Northeastern University, says he gets this question a lot. In his most recent study on the topic, he and his team analyzed more than 17,000 Android apps and did not find a single example of an app covertly activating a phone’s microphone to collect audio recordings.

While it’s theoretically possible that apps could turn on your mic, have artificial intelligence (AI) transcribe what you’re saying and convert that into data for advertisers, it would take a lot of processing power, says Choffnes. The reality is that there are far more effective ways to collect troves of data on you—and they’re much more revealing and widespread.

In fact, the Internet is virtually covered with tracking technology. As you explore, these tools may record your screen, Web searches, and interests as well as what you click on, “like,” look at, and buy.

We often inadvertently share data from GPS tracking on our phones with companies when we download and use apps, allowing them to log where we go and see who we spend time with in exchange for turn-by-turn directions and local deals.

So, say your friend told you she was off to Cancún, and now you’re getting ads for tourism packages with images of white-sand beaches. She also might have been searching for flights to Mexico while connected to your home Wi-Fi, and now her searches are connected to you—and that could lead to your having ads for a similar trip appear on your screen. (One way to prevent this is to go to myadcenter.google.com and turn off personalized ads.)

Perhaps your phone isn’t recording your voice 24/7, but Amazon’s Alexa most definitely is when you wake her up—and that intel can be used for targeted ads. “One issue here is that people may not be aware that what they say is being recorded and used for profiling and/or ad targeting,” says Choffnes. “When we think of an ‘assistant,’ we might think of a conversation with an individual trying to help us, not some system that is analyzing our voice interactions for monetization and potentially sharing this data with other parties.”

That nagging feeling that your phone is listening to you hints at a much larger phenomenon: We are being tracked by a huge data-collection network, and companies are gathering rich data on you through several means. They’re also making billions on that information—and in some cases are being targeted by hackers just as we are as individuals. Stay informed about the settings and opt-outs across all your devices to keep your data safe.

Lauren Krouse

Lauren Krouse is a journalist who covers domestic violence, trauma, and mental health. She is the author of The Ties That Harm: Understanding Sibling Abuse and Finding a Path Toward Healing. Follow her work on Substack and Instagram.