To find what exists in fact rather than artificial or fake online has become a perpetual exercise of judgment. We feel compelled to search for a shred of evidence that proves something is true or real. Yet, the vast majority of us (88%) struggle to determine what is authentic on the internet.
It’s not necessarily whether AI was used but more about whether a person, system or organization is actually trustworthy. The path of least resistance is to simply assume everything is fake and nothing can be trusted but that would be exceedingly limiting.
Technology continuously changes and the indications and signs we previously relied on may no longer be viable. What I wrote about trustworthy sources in May 2024 is already outdated. It is more about understanding the context, the source and verification of what you now encounter online.
In addition to my recent encounter with AI-generated music, a person I follow with a very large following from the original 2003 LinkedIn, wrote about a piece of writing on Substack they found inspiring. I read it but that’s not what struck me. It was that the author’s photo that was an AI-generated white-haired, blue/gray-eyed woman with Chinese characters at the bottom. I screenshot the characters and found that they were generated by Doubao AI. It’s the name of an AI chatbot and development platform by ByteDance who also owns TikTok. At 60, she refers to herself as “Anna, the Happy Senior.” The people I know who are in their 60’s do not refer to themselves as a senior.
I did a search in July if this Substack writer was a real person. Originally, it was suspect since there was lack of verification and she claimed to be a researcher at a “Marginalia Institute” which doesn’t exist. She now describes herself as “An intersection of ancient holistic wisdom and modern neuroscience. We explore healing rooted in balance. At 60, A happy life is only just beginning.” A current search returns she is a real person using the handle @slowwisdow which I assume stands for slow wisdom with an upside down “m”.
On a recent post in September, she writes about having breakfast with a colleague. The cover picture is an attractive blonde woman probably in her 40’s, not a “senior.” She mentions messages she received on her Weibo which is China’s largest microblogging and social media platform. On another recent post, the numbers corresponding to the points she makes are repetitive and duplicated and so are many of the posts themselves.
This is not the so-called 60-year-old woman presented in the photo writing this Substack. We aren’t likely to ever know who is behind this Substack persona and how much AI is used to generate the content.
You don’t have to conduct a full-on investigation like I did to prove a hunch and illustrate a point, but you do need to be cautious.
These are some questions to consider and strategies to help you verify what you encounter online is actually real and ultimately trustworthy.
The Source
Many times, it is not even a person or company name, but a nonsensical group of words or letters strung together. The photo could be pulled from a stock, digitally altered or AI-generated. I ran the photo from the suspicious Substack author through one of the reverse image searches and nothing was found, likely due to it being Chinese AI-generated.
Does the “About” section seem plausible or is it nonexistent? It’s making me rethink mine. Too little personal information may not seem real, too much becomes a privacy liability.
Does the website or account have typos, misspellings or strange domain endings like ‘.co’, ‘.online’ or ‘ltd’?
Even if the website or account looks professional with features like trust badges or verified checkmarks, it doesn’t guarantee the source is.
The Content
Does the content on the site appear repetitive, vague, with no real photos that changes often to a new forwarding address or different company name?
Is there an excessive and sketchy amount of advertising?
Is there shocking or emotional language that is intended to create animosity and assign blame? It’s what is referred to as “ragebait” to increase clicks, traffic and engagement (reactions, comments and shares).
Information manipulation is often ‘cherry-picked’ facts or mixed truth with lies to mislead you.
Social media trends are not evidence.
Verification
A quick online search can verify if a person or business is referenced anywhere else. It can also verify if an address is real, what the building looks like and a business website and phone number.
I searched my own name and within seconds returned Whitepages.com revealing an old landline phone number, a blocked-out cell phone number and my current as well as previous addresses. Also, a local newspaper verifies the same address along with name and when the property was purchased, all public information.
If the “About” section or content lists companies or institutions, look them up to verify they even exist and appear to be legitimate.
Extraordinary claims require strong evidence. Always verify with reputable sources.
Cross-check information especially for urgent or emotional requests for money.
Verify unexpected payment changes independently and call official phone numbers rather than trusting caller ID or links in messages.
Use the SIFT method: Stop, Investigate the source, Find other trusted coverage and Trace claims back to their original, authentic context. (University of Rochester)
We are all worried about scams, about a family member falling for something or whether something is being used to deceive us. However, we shouldn’t negate everything online as fake or harmful. The music and Substack examples are fairly benign and inconsequential, but they are great ways to exercise and strengthen your evidence-gathering skills before you encounter something more dire.
Good, trustworthy people, institutions, companies and resources do exist online. You just have to find and verify them.
The rise of AI suspicion | ASU News