Seeing Is No Longer Believing: How to Verify Online Identities in the Age of AI

Facts routinely challenge appearances. Artificial intelligence has made careful verification more important.

We used to tell people to look for obvious warning signs when interacting with strangers online. Poor grammar. Suspicious photographs. Inconsistent stories. Refusal to participate in a video call. Those were reasonable precautions.

Today?

The grammar is impeccable. The photographs are beautiful. The professional history is convincing. The LinkedIn profile has hundreds of connections. That friendly, articulate, seemingly successful individual smiling at you during a live video call? They may not exist.

Welcome to the age of artificial intelligence, where seeing is no longer necessarily believing. Knowing the difference can protect far more than your bank account.

A Convincing Profile Is Not Independent Proof of Identity.

Let’s begin with something rather obvious that we seem to have collectively forgotten.

A LinkedIn profile is not an identity document. Neither is a Facebook account, an Instagram page, a professional-looking website, or an impressive biography. They are carefully crafted representations.

Representations can be manufactured. With readily available technology, someone can create a credible professional identity complete with photographs, employment history, educational credentials, business affiliations, and a seemingly established online presence.

Now, give that identity some time, a few connections, and regular activity, and it begins to look remarkably legitimate.

The more sophisticated operators don’t necessarily create everything from scratch. Sometimes they impersonate real people. A legitimate executive. A successful investor. A recruiter at a respected company. A financial advisor with genuine credentials. The person exists. The company exists. The credentials may even check out.

The problem is that the person communicating with you may have absolutely nothing to do with any of them. And that’s a very different problem from determining whether a name appears in a database.

“But I’ve Had a Video Call With Them!”

This is where things get particularly interesting.

For years, a live video call was considered one of the better ways to establish whether an online acquaintance was genuine. After all, you could see their face. Hear their voice. Watch them respond to questions. Surely that meant something.

Not necessarily anymore.

Altered video and voice spoofing can make online impersonation more convincing. The FBI has reported complaints involving deepfakes and stolen personal information in remote-work applications, including suspected voice manipulation during online interviews. These are reports of suspected techniques, not a claim that every convincing call uses them. FBI advisory

A video call may still provide useful information, but it should not be treated as independent proof of identity.

In a case reported to Hong Kong police in January 2024, an employee authorized transfers totaling about HK$200 million after a fabricated video conference impersonating a senior company officer. In its June 2024 account, the government said investigators believed the conference had been prerecorded using publicly available video and voice material; there was no actual interaction during the meeting. Hong Kong Government account

This case illustrates deception using a prerecorded conference, not proof of a live interactive deepfake.

Think about that.

The fraud didn’t succeed despite the video call.

The video call was part of the deception.

Technology has become sophisticated enough that the very tools we once relied upon for reassurance can now be used to manufacture it.

A Real Person. A Real Company. A Completely Fabricated Relationship.

The distinction matters:

Verifying that someone exists is not the same as verifying that you’re communicating with that person.

Suppose someone contacts you on LinkedIn. They claim to be a managing director at a well-known investment firm. You search the firm. It’s legitimate. You find the executive’s biography. It’s legitimate.

You confirm the person’s professional credentials. Also legitimate. You may even find interviews, conference appearances, and published articles. Everything checks out.

Except one rather important detail. You haven’t established that the individual contacting you is actually that executive. You’ve verified the identity being presented. You haven’t authenticated the person presenting it.

A Hypothetical Remote-Hiring Example

Hypothetical example: A remote applicant presents another professional's genuine credentials. Confirming those credentials establishes that the professional exists; it does not establish that the applicant is that person. Before equipment, access, or sensitive information is released, the employer needs an appropriate identity-verification process as well as a background review.

That gap is where impersonation fraud thrives. The more impressive the identity, the easier it can be to overlook the need to authenticate the person presenting it.

Fraudsters Have Learned Something About Human Nature

People like to believe they can recognize deception. We trust our instincts. We read facial expressions. We evaluate tone, confidence, and credibility. Most of us believe we’re reasonably good judges of character.

Unfortunately, sophisticated fraudsters understand that. They don’t necessarily begin with a request for money.

They begin with credibility. A professional introduction. A shared interest. A mutual connection. An interesting business opportunity. Perhaps several weeks of perfectly ordinary conversation. Maybe even months.

They learn how you communicate. What matters to you. What you’re working on. Who you know. They establish familiarity. That familiarity has a remarkable ability to lower our defenses.

By the time an investment opportunity, financial request, or sensitive document enters the conversation, the individual may no longer feel like a stranger.

They feel like someone you know.

But familiarity is not verification. And time spent communicating is not evidence of identity.

According to the Federal Trade Commission's April 2026 reporting, consumers reported $2.1 billion in losses from scams originating on social media during 2025. This is reported social-media-scam loss, not a measure of AI or deepfake losses. FTC data spotlight

That figure reflects reported losses, not necessarily the full extent of the problem.

And the financial damage is only part of the story.

Businesses are compromised, personal information is exposed, reputations are damaged, and individuals discover that a relationship they believed was genuine was carefully manufactured.

Five Checks Before Trusting an Online Introduction

Not every new LinkedIn connection requires an investigation. But when a relationship moves toward money, confidential information, professional access, or significant personal commitments, establish a few things independently.

1. Verify the person through a channel they did not provide.

If someone claims to represent a company, locate the company’s official contact information yourself. Call the main office. Contact an independently verified corporate email address. Ask to be connected with the individual. Do not rely solely on a telephone number, email address, or website provided by the person you’re investigating. Remember, the entire point is to establish an independent source.

1. Verify the credentials, not merely the biography.

Professional licenses, business registrations, and regulatory records can provide important information. For financial professionals, FINRA BrokerCheck is a useful starting point. For attorneys, accountants, and other licensed professionals, consult the appropriate licensing authority. Verifying a credential does not prove that the person communicating with you owns it.

1. Look for a history that exists outside their control.

Legitimate professional identities often leave records across multiple independent sources over time. Corporate filings. Professional directories. Historical business affiliations. Licensing records. Court records, where relevant. No single source establishes everything. The question is whether the information forms a consistent, independently supportable history.

1. Don’t let a video call become your final test.

A live conversation may reveal inconsistencies, but visual recognition and a familiar voice are not conclusive authentication. If the relationship involves a financial transaction, independently confirm the recipient and account details through a trusted contact method already on file. This precaution does not depend on proving which technology an impersonator used.

1. Pay attention when someone attempts to control the verification process.

Are they directing you exclusively to websites they supplied? Insisting that you communicate only through a particular application? Discouraging you from contacting their company independently? Creating urgency around an investment or transfer? Becoming offended when you ask reasonable questions? None of these behaviors automatically establishes fraud. But they may justify stopping until the discrepancies are resolved. The person asking for your trust should not also control every source you use to verify them.

When the Stakes Increase, So Should the Due Diligence

There is a significant difference between accepting a LinkedIn connection and entering into a business relationship with someone you’ve never met.

If you’re considering an investment, business partnership, private lending arrangement, significant financial transfer, or another consequential commitment, a quick internet search may be woefully inadequate.

That’s where professional due diligence becomes valuable.

At Axiom Verify, we examine identities, professional histories, business affiliations, financial representations, litigation, and other relevant information using lawfully accessible records and investigative resources.

Our work can extend across jurisdictions and international borders. We examine whether the information presented is consistent with independently obtained evidence. We look for relationships between individuals, businesses, and financial interests.

We identify discrepancies, historical patterns, and areas requiring further examination. We document our findings with sources.

An important limitation to understand: a background investigation can establish whether records support a claimed identity and history, but it does not, by itself, authenticate the person currently operating an online account or appearing on a video call.

That may require additional identity authentication through trusted channels. The objective is to build confidence through independent evidence, not through the sophistication of someone’s presentation.

The Most Expensive Mistake May Be the Assumption You Never Questioned

Consequential decisions often turn on surprisingly small assumptions. We assume the person is who they claim to be. We assume the company affiliation is legitimate. We assume the credentials belong to the individual contacting us. We assume the investment exists as represented. We assume the face on the screen belongs to the person speaking.

Sometimes, every one of those assumptions is correct. But when they’re not, the consequences can be extraordinary. The challenge isn’t that technology has made verification impossible.

It hasn’t.

The challenge is that technology has made the appearance of credibility easier to manufacture. Which means our standards for establishing credibility must evolve.

Quick Answers

Can someone use AI to impersonate another person during a live video call?

Manipulated video and voice spoofing can be used in online impersonation. The FBI's remote-work advisory describes reported suspected voice manipulation during interviews. A call may be useful as one component of verification, but it is not definitive proof of identity. The Hong Kong case described above involved a prerecorded conference, not an interactive live deepfake.

Can a reverse image search detect an AI-generated profile photograph?

Sometimes it can identify stolen or reused photographs. But a newly generated image may have no prior online history. An unsuccessful reverse image search does not establish authenticity.

How can I verify a recruiter who contacts me on LinkedIn?

Contact the employer through its independently located website or main telephone number. Confirm that the recruiter is authorized to represent the company and that the position exists. Be particularly cautious of requests for payments, banking information, or sensitive identity documents before a legitimate hiring process has been established.

What if the person becomes offended when I ask to verify their identity?

Their reaction is not proof of wrongdoing. But reasonable verification is an ordinary part of responsible business and financial decision-making. You are entitled to establish who you’re dealing with before exposing yourself to risk.

Can Axiom Verify determine whether someone is using a false identity?

Our investigations can identify inconsistencies between a claimed identity and independently obtained records, uncover relevant affiliations, and document evidence of possible misrepresentation. The ability to establish a definitive identity depends on the records, authorizations, and authentication methods available in the particular case.

What should I do if I suspect I’ve been deceived?

Stop sending money or sensitive information. Preserve communications, transaction details, and identifying information. Contact your financial institution immediately if funds have been transferred. Report suspected fraud to the relevant platform and appropriate authorities, including the FTC fraud-reporting service.

One Final Thought

Artificial intelligence is creating remarkable opportunities in medicine, business, research, communication, and countless other fields.

It is also changing the economics of deception.

What once required substantial time, specialized skills, and elaborate preparation can increasingly be accomplished with readily available tools.

The answer is not to become suspicious of everyone. The answer is to be more disciplined about what counts as evidence.

A photograph is not proof. A profile is not proof. A familiar voice is not proof. And increasingly, even a live video call is not proof.

Trust is important. But trust and verification are not mutually exclusive.

At Axiom Verify, we don’t begin an investigation assuming someone is dishonest. We begin with questions. Then we follow the evidence. Because in a world where almost anything can be convincingly manufactured, the ability to distinguish appearance from fact has never been more valuable.

We don’t know whether someone is misrepresenting who they are. Neither do you. That’s why we verify. Before entering into a consequential business, financial, or personal relationship, know who you’re dealing with.

When you have to know—know for Certain.

Related Reading

See due diligence background checks and affinity fraud in professional networks for complementary questions about history and relationships.

Sources

• Hong Kong Government: June 26, 2024 account

• FTC: reported 2025 social-media-scam losses

• FBI: deepfakes and stolen identity in remote-work applications