Digital Payments in Iraq: A Security Checklist
How to accept wallet, QR, card and online payments in Iraq safely: server-side confirmation, webhook checks, reconciliation and account protection.
Iraq's SIM re-registration, WhatsApp BSUIDs and NIST's new SMS rules all change OTP. How to stop treating a phone number as an identity.

Many apps built in Iraq over the past decade quietly assume that a phone number is a person. Sign-up asks for a number, an OTP confirms it, and from then on that number is the account. Three changes in 2026 make that assumption harder to defend. I think every team that relies on phone number identity should review its design now, rather than after the next account takeover.
First, regulation in Iraq. The Communications and Media Commission approved new SIM rules under Board of Commissioners Decision No. 116 on 29 September 2026. According to 964media's report:
Second, WhatsApp. Optional usernames are rolling out, so people can message a business without sharing their phone number. Instead, Meta provides a business-scoped user ID (BSUID), and the number may stop appearing in webhooks.
Third, standards. NIST SP 800-63B-4 now classes SMS and phone-network one-time codes as a "restricted" authenticator.
Each change is reasonable on its own. Together, they tell us that a phone number is a contact channel with a lifecycle, not a permanent identity.
Registration checks tell you who bought a line, not who is holding the phone today. An INSM paper on SIM registration in the Kurdistan Region makes exactly this point. Sellers already ask for ID, a residence card, a live photo and a fingerprint, yet those checks describe the registrant, not the daily user.
The same paper notes that in August 2026 the CMC temporarily limited SIM sales in the Kurdistan Region to operators' main centres. This followed complaints that lines had been registered in citizens' names without their consent.
I see the regulator's tighter checks as a positive step. But developers should take the INSM paper's wider argument to heart: protection must cover a number's whole life. That means sale, sharing, transfer, inactivity, recycling and removal from digital accounts.
Your app sits at the end of that chain. If a number is recycled or re-registered to someone else during the 720-day period, an account keyed only on that number can pass to a stranger with one SMS.
A BSUID is persistent, but it is unique to each pairing of business portfolio and user. The same person therefore has a different BSUID with every business. It cannot serve as a cross-business key, and you should not try to use it as one.
The detail that matters most for OTP is that authentication templates cannot be sent to a BSUID. The 8x8 release notes on BSUIDs map Meta error 131062 to the instruction to send OTP or authentication templates to the phone number instead. So even if a customer chats with you through a username, you still need a verified phone number on file for authentication.
In practice, your webhook handler must accept messages that contain a BSUID and no phone number. It should link them to an existing user where possible, and never create a duplicate account just because the number field is empty.
There is also a client-side deadline that has now passed. Meta deprecated the old PendingIntent handshake for zero-tap code delivery on Android on 15 October 2026. Its zero-tap authentication templates page recommends moving to the OTP Android SDK.
The handshake now includes a request_id (a UUID) that your app must store and check when the code arrives. Zero-tap works only on Android; other devices see a copy-code button. These templates do not support URLs, media or emojis.
According to TypingDNA's summary of SP 800-63B Rev 4, this is the first time NIST has used the "restricted" category. Restricted does not mean banned. An organisation that keeps SMS OTP should have a documented risk assessment, a migration plan and a notice to users.
One-time codes can be phished, so they never reach NIST's highest assurance level (AAL3). Limits on repeated failed attempts are expected. Codes should be at least six digits and accepted only once.
If you run an ISO/IEC 27001 management system, this fits neatly into your risk register. Record:
The core fix is simple. Every user gets an internal ID, and phone numbers and BSUIDs become attributes that can change. Here is an example Laravel migration:
// Example migration: the phone number is an attribute, not the key
Schema::create('user_contacts', function (Blueprint $table) {
$table->id();
$table->foreignId('user_id')->constrained()->cascadeOnDelete();
$table->string('phone_e164', 20)->nullable()->index();
$table->string('wa_bsuid', 140)->nullable()->index(); // scoped to your business portfolio
$table->timestamp('phone_verified_at')->nullable();
$table->timestamp('last_active_at')->nullable();
$table->timestamps();
});
And an example rate limiter for OTP sends. The limits are illustrations; tune them to your traffic.
// Example: in AppServiceProvider
use Illuminate\Cache\RateLimiting\Limit;
use Illuminate\Http\Request;
use Illuminate\Support\Facades\RateLimiter;
public function boot(): void
{
RateLimiter::for('otp-send', function (Request $request) {
$phone = (string) $request->input('phone');
$range = substr($phone, 0, 9); // groups sequential numbers
return [
Limit::perMinute(1)->by('phone:'.$phone),
Limit::perHour(5)->by('phone:'.$phone),
Limit::perHour(30)->by('range:'.$range),
Limit::perHour(10)->by('ip:'.$request->ip()),
];
});
}
// routes/api.php
Route::post('/otp/send', [OtpController::class, 'send'])
->middleware('throttle:otp-send');
The range limit matters because of SMS pumping. Bots request OTP messages to number ranges where a rogue carrier shares the revenue, and you pay for every message. The usual warning signs are:
At BulkSMSIraq, which I founded, we deliver OTPs over WhatsApp, Telegram and SMS from one API, with automatic SMS fallback when WhatsApp or Telegram fails. That is one way to handle the sending side. The identity model and rate limits above still belong in your own application, whichever provider you use.
Not every action needs the same proof. This is the table I suggest teams start from:
| Action | Suggested method | Notes |
|---|---|---|
| Low-risk sign-in, notifications | WhatsApp or SMS OTP | SMS as fallback when WhatsApp is unavailable |
| Phone number change | OTP to new number plus a second check | Notify the old number and email |
| Account recovery | OTP plus a further factor or identity check | Never one SMS alone |
| Admin, payments, payout changes | Passkeys | Phishing-resistant |
Passkeys are no longer niche. The FIDO Alliance's October 2026 Passkey Index estimates 7 billion passkeys in use worldwide. For high-value actions, they are the direction I would plan towards.
request_id.Your team may need help with the Cloud API side, such as BSUID handling or webhooks. My company, Standing, is an official Meta Tech Partner for WhatsApp and builds WhatsApp Business messaging and chatbots, so we can help.
The takeaway is short: treat a phone number as a channel you verify and re-verify, not as the person. Build the account around your own ID and match the proof to the risk. Then the changes coming from regulators, Meta and NIST become routine maintenance rather than a crisis.
🤖 Researched and drafted with AI assistance, guided by my editorial standards. Spotted something wrong? Tell me.