Methods.

At first, the only thing you actually know is that you are communicating with someone who controls an account.

Everything beyond that is evidence to be accumulated.

A convincing story is still a story.

Not all evidence is equal.

A text claim is weak.

A photo is stronger, but can be stolen.

A prerecorded video proves someone possesses that video.

A public social-media account helps, but accounts can be copied or impersonated.

Consistent information across unrelated sources is much stronger.

Repeated live interaction containing unpredictable conversation is stronger still.

Independent real-world corroboration is stronger again.

The basic principle is:

Possession of evidence is not necessarily proof of identity.

If someone is becoming emotionally important, financially involved, or otherwise consequential in your life, eventually there should be normal live interaction.

A video call does not need to become a hostage-proof-of-life ceremony.

Just talk.

Unexpected questions, reactions to things happening in the room, continuity across conversations, and ordinary spontaneous interaction establish far more than dozens of photos ever will.

Declining a video call once proves almost nothing.

Indefinitely preventing every reasonable means of live verification is different.

Certain stories naturally explain communication difficulties:

• military deployments
• classified workplaces
• oil rigs
• international contracting
• medical work
• remote travel
• war zones
• restricted facilities

Some of those situations are completely real.

But an explanation for why verification is difficult does not itself become verification.

The correct conclusion is not necessarily:

“They are lying.”

It is:

“This remains unverified.”

That distinction prevents a lot of bad reasoning.

One inconsistency is usually noise.

Several independent inconsistencies are a pattern.

If important details begin conflicting, write them down.

Then ask a very simple question:

Which explanation requires fewer assumptions?

World A: the person's account is substantially true.

World B: some important part of the identity or story is fabricated.

If World A eventually requires six unrelated coincidences and elaborate explanations while World B requires one assumption, that is useful information.

You do not need courtroom-level proof before reducing your exposure.

If something feels wrong, check public information.

Search usernames.

Search names.

Reverse-search important images.

Compare timelines.

Look for duplicated biographies, conflicting identities, impossible location overlaps, or accounts claiming years of history that appeared three weeks ago.

The goal is authentication, not stalking.

Do not break into accounts, harass family members, dox people, or turn an online conversation into amateur intelligence work.

You only need enough information to decide how much trust is justified.

Once money enters the conversation, standards should become much stricter.

Do not send money to someone whose identity and circumstances you cannot independently establish.

That includes:

• loans
• gift cards
• cryptocurrency
• investment opportunities
• emergency expenses
• travel costs
• shipping fees
• customs fees
• medical expenses
• phone bills
• military leave fees
• receiving money on someone's behalf

And never allow someone you barely know to use your bank account, identity documents, mailing address, or financial infrastructure.

One rule is particularly useful:

Never spend money to solve uncertainty about someone's identity.

The response itself contains information.

A reasonable person may be surprised or mildly annoyed.

But they can usually understand something like:

“Nothing personal. I am careful online.”

A very different response is an attempt to turn verification into a loyalty test:

“You don't trust me?”

“After everything I've told you?”

“If you cared about me you wouldn't ask.”

“Why would I ever lie to you?”

None of those statements provide evidence.

They change the subject from identity to emotion.

Notice that.

It helps to stop thinking of people as either “trusted” or “not trusted.”

Think in stages.

Stranger

Conversation only.

Consistent stranger

Their story has remained coherent over time.

Live verified

You have had repeated normal live interaction.

Identity corroborated

Independent evidence supports important claims.

Real-world known

You have meaningful interaction across multiple channels or have met safely in person.

Trusted

Trust exists because of substantial history.

And one distinction matters enormously:

Authentication proves who someone is. It does not prove they are trustworthy.

Real people scam people too.

This may be the hardest one for analytical people.

An inconsistent story can become interesting.

Interesting becomes a puzzle.

And puzzles attract attention.

But you are not required to solve another human being.

When something stops making sense, the question should become:

Has this person earned the next increment of trust?

If the answer is no, simply stop increasing exposure.

You can remain polite.

You can remain curious.

You can even keep talking.

But money, private information, emotional commitments, and serious decisions wait for evidence.