Who Are You Really Talking to Online? A Guide to Social Media Safety

Social media platforms like Facebook can be wonderful tools. They help us stay connected with family who live far away, share photos of our grandkids, and join communities of people who share our hobbies. As we've heard from our readers, many of you use them every day to keep in touch.

But alongside the joy of connection, these platforms can also bring a sense of unease. Who is seeing our information? Are the people we interact with always who they say they are? These are smart questions to ask. Today, let’s take a look at navigating the world of social media safely, with a focus on protecting your privacy and spotting the kinds of fake accounts that are becoming more common.

What do you think about using Facebook?

Social Media, Safely: Your Guide to a More Secure Experience

A Word on Privacy: The Meta AI Example

It's no secret that social media companies use our data. Sometimes, however, they roll out new features without making it entirely clear how our information is being used, which can lead to privacy problems.

• A Recent Real-World Example: Recently, Meta (the company that owns Facebook and Instagram) launched a new AI app. Many users didn't realize that their private questions to the AI assistant could be publicly shared in a "Discover" feed if they hit a "Share" button. This led to personal and sensitive conversations being made public unintentionally. While Meta has since added warnings, it’s a powerful reminder that we need to be mindful and deliberate about what we click and share.

• Read more about this here: https://techcrunch.com/2025/06/12/the-meta-ai-app-is-a-privacy-disaster/

The Rise of Fakes: Bots, AI Profiles, and What to Look For

As technology gets smarter, so do the methods for creating fake accounts and content. Here’s what to be aware of:

• What are "Bots"? In this context, a bot is not a physical robot. It is an automated software program designed to act like a human on social media. These bots can be programmed to automatically post comments, "like" content, and follow thousands of accounts to make a person or idea seem more popular than it really is. People, companies, and governments pay for these services.

• How AI is Making It Worse: Today, AI can create entire fake profiles from scratch, complete with a highly realistic but completely fake profile picture (often called a "deepfake"). It can also generate generic-sounding comments and posts, making it harder than ever to tell if you're interacting with a real person or an automated account.

This is what a bot farm looks like! It’s basically a computer running a complicated program hooked up to a bunch of phones:

How to Spot the Fakes: Your Toolkit for Being a Savvy User

The goal is not to be paranoid, but to be a discerning observer. Here are some simple tips to help you spot accounts that might not be what they seem:

1. Examine the Profile Picture: AI-generated faces can look incredibly real, but they often have small, tell-tale glitches. Zoom in and look for odd details. Are the earrings mismatched? Are the pupils of the eyes different sizes? Is the background strangely blurry or distorted? If there’s any text in the image see if any of the letters are malformed or the words don’t make sense. Sometimes a person will look plastic-y or have an unrealistic sheen. I have to admit, it is getting harder and harder to tell!

2. Check Their Activity: Bots often have strange activity patterns. They might have a very low follower count but be following thousands of other accounts. Or they might post dozens of times in a very short period, or make lots of vague, repetitive comments like "Great post!" on many different pages.

3. Look at Their "Voice": Read their posts or comments. Does the language feel a bit flat, generic, or lacking in personal stories and a unique point of view? This can sometimes be a sign of AI-generated text.

4. Trust Your Gut: If an account, a friend request from someone you don't know, or a message just feels "off," it's okay to ignore it. As we've heard from readers, there is a general concern about scams, and trusting your instincts is your best first line of defense.

AI-generated video has gotten better as well. Using these tips, try out this quiz and see if you can spot the AI videos:

https://www.nytimes.com/interactive/2025/06/29/business/ai-video-deepfake-google-veo-3-quiz.html

I went 10/10, but it’s getting much harder!

Quick Tech Tip: Do a Reverse Image Search

If you're suspicious of a profile picture, you can often check its origin. Save the profile picture to your device. Then, go to a search engine like images.google.com. You'll see a small camera icon in the search bar. Tap it and upload the photo you just saved. The search results will show you where else that image has appeared online. If it's a common stock photo or appears on many other random websites, it's a strong sign that the profile is not genuine.

Tech Term Demystified: Deepfake

A "Deepfake" is a term for a highly realistic but completely fabricated image, video, or audio recording created using Artificial Intelligence. The "deep" part refers to "deep learning," a type of AI that is exceptionally good at learning and mimicking patterns. This technology can be used to swap one person's face onto another's in a video or to create an entirely new, photorealistic image of a person who doesn't even exist. While it can be used for entertainment, it is also a powerful tool for creating misinformation, making it more important than ever to be critical of the media we see online.

Good News Byte

In response to the challenges of deepfakes and AI-generated content, many technology companies and research institutions are actively working on new "AI detection" tools. These are sophisticated programs designed to analyze an image or video and determine the statistical likelihood that it was created by an AI. While the technology is still evolving, this represents a major effort to create tools that can help all of us more easily distinguish between what's real and what's fake.

Did You Know?

The concept of automated "bots" that trick people goes back to the very beginning of computing. In 1950, the famous computer scientist Alan Turing proposed what is now known as the "Turing Test." The test was a simple game: could a human, by having a text-based conversation, reliably tell the difference between another human and a computer program that was trying to imitate a human? This fundamental question about artificial communication is more relevant today than ever before. Most modern AI can pass the “Turning Test” with flying colors.

Your Turn to Be a Digital Detective!

This week, when you're browsing online, pay closer attention to the profiles you see. Without clicking on anything suspicious, just practice observing. Look at the profile pictures and the way people comment. See if you can spot any of the tell-tale signs of an automated or fake account. It's a great way to sharpen your digital literacy skills!

Wishing you a safe and authentic week online!

Steve