elameen.php

Building @Elv_HQ & @ElvLab | A multi-service consultancy for coding, fintech & research

[elameen.php @elvmeen]

My X account is 10 years old and still not growing. After studying X’s open-sourced “For You” algorithm thoroughly, I built an automation to post 500 tweets over the next month & see what actually goes viral.

By the end, I’ll have learned something, or I’m deleting this account.

[elameen.php @elvmeen]

Cybersecurity women are not consistent. They're mostly gays.

[korz3yn @webenstein_]

Can a cybersecurity guy date a cybersecurity girl? Asking for a friend.

[elameen.php @elvmeen]

There is something deeply unserious about calling a system “zero trust” while every internal application trusts the same service account.

[haffab @thsblackkk]

Real intimacy start by exchanging playlist that's one of my love language

[elameen.php @elvmeen]

@RedHatPentester Lol. I, that almost deleted Twitter. 🤣🤧

Glad you find my content meaningful. I’m a Cybercrime/Cybersecurity researcher with several publications in some top tier international journals. I’m open to any possible collaborations.

More about me here: https://t.co/lJBuCWyDGn

[elameen.php @elvmeen]

Ransomware payment decisions are not simply “pay” versus “don't pay.”

Once data exfiltration enters the equation, organisations are negotiating about availability and disclosure.

Would you still support an absolute “never pay” policy?

[elameen.php @elvmeen]

A company can survive an attack and still fail the victims.

Recovery metrics should include people, not just uptime.

[elameen.php @elvmeen]

The best cybersecurity policy is useless if nobody can follow it during an actual incident.

Write for panic, not peace.

Incident responders: what instruction do people forget first when things go wrong?

[elameen.php @elvmeen]

I like AI.

That is exactly why I don't trust every ridiculous claim made about AI.

[elameen.php @elvmeen]

African cybersecurity research shouldn't exist only as “apply Western model to African country.”

There are local realities worthy of theory, not merely replication.

What African security problem deserves its own model?

[elameen.php @elvmeen]

People complain that users don't read security warnings.

Have you read security warnings?

Half of them look like a lawyer arguing with a toaster.

[elameen.php @elvmeen]

One reason social engineering remains effective is simple:

The attacker can choose when to attack.

The victim does not choose when to be tired.

[elameen.php @elvmeen]

I don't think cybersecurity needs more fear.

It needs better explanations.

Fear gets attention.

Understanding changes behaviour.

[elameen.php @elvmeen]

The internet has trained everyone to expect sophisticated hacks.

Meanwhile someone has left .git exposed in production again.

[elameen.php @elvmeen]

AI ethics without AI security feels incomplete.

A system cannot be meaningfully trustworthy if an attacker can manipulate what it does.

Agree or are these separate enough to study independently?

[elameen.php @elvmeen]

Research becomes better when we stop treating replication as an insult.

If a finding only works once, in one population, under one exact setting, perhaps we should know that.

Why is replication still treated as academically boring?

[elameen.php @elvmeen]

Some people read “open source” and immediately interpret it as “secure.”

My brother, the vulnerabilities are also open source.

[elameen.php @elvmeen]

Cybercriminals do not need victims to be stupid.

They need victims to be human at the correct moment.

[elameen.php @elvmeen]

There are organisations where the cybersecurity policy is 70 pages long and the actual password is still Company@123.

Policy achieved.

[elameen.php @elvmeen]

Humans are not security bugs.

They are part of the system.

Design accordingly.

[elameen.php @elvmeen]

I have seen enough AI demos to develop a simple rule:

If the demo requires perfect prompts, perfect context and perfect conditions, I am watching theatre.

What's the most misleading AI demo pattern you've noticed?

[elameen.php @elvmeen]

Cybersecurity prediction research has a weird obsession with asking:

“Can the model predict this dataset?”

I care more about:

“Can this survive contact with a real network?”