Dalil App 5 Million Users Exposed via Open MongoDB

2019 · 5M users affected

By Karim El Labban · ZERO|TOLERANCE

In 2019, security researchers discovered that Dalil, a popular Saudi Arabian caller

identification application with over 5 million users, had left its entire MongoDB

database exposed to the internet without any authentication or access controls. The

585-gigabyte database contained detailed personal information including real names,

phone numbers, precise GPS location data, device information, and mobile carrier

details for millions of Saudi users.

The database had been publicly accessible for an undetermined period before its

discovery, during which time any individual with basic technical knowledge could have

accessed, downloaded, or manipulated the complete user dataset.

## Key Facts

  • .**What:** Dalil caller ID app left 585GB MongoDB database open without authentication.
  • .**Who:** Over 5 million Saudi users of the Dalil app.
  • .**Data Exposed:** Real names, phone numbers, GPS locations, device IDs, and carrier data.
  • .**Outcome:** Pre-PDPL; zero security controls on a surveillance-grade dataset.

## What Was Exposed

  • .Real names associated with phone numbers for approximately 5 million users,

predominantly Saudi Arabian nationals

  • .Phone numbers including both the primary number and contacts from users' address

books uploaded by the app

  • .Precise GPS location data with latitude and longitude coordinates, revealing

users' physical movements and frequented locations

  • .Device information including phone model, operating system version, IMEI numbers,

and device identifiers

  • .Mobile carrier information revealing which telecommunications provider each user

subscribed to

  • .App usage data including installation dates, last active timestamps, and feature

usage patterns

  • .Email addresses associated with user accounts

The Dalil breach is a case study in the risks posed by the proliferation of mobile

applications that collect far more data than their core functionality requires. A

caller identification app needs access to a phone number database and, arguably, a

user's own phone number. It does not need continuous GPS location tracking, device

IMEI numbers, or carrier information. The collection of this excessive data, combined

with the absence of any database security whatsoever, created a surveillance-grade

dataset on 5 million people that was freely accessible to anyone who knew where to

look.

The GPS location data is the most concerning element of the exposure. With latitude

and longitude coordinates associated with identified individuals, the dataset could

be used to track users' movements, identify their home and workplace addresses,

determine their daily routines, and establish patterns of association between

individuals who frequent the same locations. In a region where personal privacy is

culturally valued and where certain gatherings or associations may carry social or

legal sensitivity, the exposure of granular location data poses risks that extend

far beyond conventional identity theft.

The data could be weaponized for stalking, blackmail, social engineering, or targeted

surveillance. A domestic abuser could track a victim's movements. A business rival

could monitor a competitor's meetings and associations. A criminal organization could

identify high-value targets based on their home addresses and daily patterns. The

potential for harm from location data is limited only by the imagination and intent

of those who access it.

The 585GB volume of the database suggests that the exposure included not only current

user data but potentially historical location data spanning the application's entire

operational lifetime. This longitudinal location dataset would enable the construction

of detailed movement profiles showing how individuals' patterns changed over time,

where they traveled, and whom they met. For security services, intelligence agencies,

or malicious actors, this type of historical location data is extraordinarily

valuable and would normally require significant resources and legal authority to

obtain.

Open MongoDB instances have been a persistent vulnerability pattern across the

technology industry, with search engines like Shodan making it trivial to discover

unprotected databases. The Dalil case demonstrates that this vulnerability is not

limited to obscure startups in mature technology markets; it is a global problem

that affects applications serving populations in every region. The fact that an app

serving 5 million users in a high-profile market like Saudi Arabia could operate

with zero database authentication suggests a fundamental lack of security awareness

and resources during the application's development and deployment.

## Regulatory Analysis

The Dalil breach occurred in 2019, four years before the enactment of Saudi Arabia's

PDPL in September 2023. At the time of the breach, Saudi Arabia lacked a comprehensive

data protection law, which meant there was no dedicated regulatory body to investigate

the breach, no mandatory notification requirements, and no specific penalties for the

type of data exposure that occurred. This regulatory vacuum illustrates precisely why

the PDPL was necessary, as the Dalil incident, and others like it, demonstrated that

the Saudi digital ecosystem had outpaced the legal framework designed to protect it.

Analyzed under the PDPL as it now stands, the Dalil breach would trigger multiple

violations. Article 6 establishes consent requirements for the collection of personal

data, mandating that individuals be informed of the purpose for which their data is

being collected and that consent be specific, informed, and freely given. The

collection of GPS location data and device identifiers by a caller ID application

raises serious questions about whether users were adequately informed about the scope

of data collection and whether their consent extended to continuous location tracking.

If the app's privacy policy did not explicitly disclose the collection of GPS data,

or if consent was bundled into a general terms-of-service acceptance, the processing

would lack a valid legal basis under Article 6. The PDPL requires granular consent

that is specific to each purpose of processing, not blanket authorization hidden in

dense legal language that no reasonable user would read or understand.

Article 10 addresses the processing of personal data by third parties and the

obligations of data controllers to ensure processor compliance. If Dalil used

third-party cloud hosting services for its MongoDB instance, the responsibility for

securing the database would remain with Dalil as the data controller. The fact that

the database was deployed without authentication suggests either a catastrophic

misconfiguration or a deliberate decision to forgo security controls, neither of

which would constitute a valid defense under the PDPL.

Article 14's requirement for appropriate technical and organizational measures is

where the PDPL would apply most directly to the Dalil case. An exposed MongoDB

database with no authentication represents the most basic possible failure of

technical security measures. There were no access controls, no encryption, no

monitoring, and no alerts. This is not a case where sophisticated attackers defeated

well-designed defenses; it is a case where no defenses existed. Under the PDPL,

SDAIA would be justified in imposing the maximum fine of SAR 5 million, as it would

be difficult to identify a more complete failure of the duty to implement appropriate

security measures.

## What Should Have Been Done

The most fundamental control that should have been in place was database

authentication and access control. MongoDB supports robust authentication mechanisms

including SCRAM-SHA-256, X.509 certificates, and LDAP integration. Any of these

mechanisms, even the simplest username-and-password authentication, would have

prevented the database from being publicly accessible. The fact that the database

was deployed without any authentication represents a failure of the most basic

security hygiene.

Network-level access controls should have restricted database connections to only

the application servers that needed access, using firewall rules, VPN requirements,

or cloud security groups to ensure that the database was never directly reachable

from the internet. Defense in depth means that even if database authentication were

somehow bypassed, network-level controls would prevent unauthorized connections from

reaching the database at all.

Data minimization should have been a core design principle from the application's

inception. A caller identification app does not need to collect continuous GPS

location data, device IMEI numbers, or carrier information to perform its primary

function. The PDPL's data minimization principle, which requires that data collection

be limited to what is necessary for the stated purpose, would have prevented the

creation of this surveillance-grade dataset in the first place. By collecting only

the data necessary for caller identification, the blast radius of any potential

breach would have been limited to phone numbers and associated names, rather than a

comprehensive profile including physical location histories.

Encryption of data at rest should have been implemented to protect the database

contents even in the event of unauthorized access. MongoDB Enterprise and even the

community edition support encryption at rest using industry-standard algorithms. If

the database had been encrypted, an attacker who gained access to the raw files or

the database connection would have been unable to read the data without the encryption

keys. Combined with column-level encryption for the most sensitive fields such as GPS

coordinates and national identification numbers, encryption would have provided a

critical last line of defense.

Continuous security monitoring and vulnerability scanning should have been part of

the application's operational procedures. Automated tools that scan for exposed

databases, open ports, and misconfigured cloud services are widely available and are

considered a baseline requirement for any internet-facing application. Regular

penetration testing would have identified the open MongoDB instance as a critical

vulnerability, and a vulnerability management program would have ensured that the

issue was remediated promptly. The extended period of exposure suggests that no such

monitoring or testing was in place, leaving the 5 million users' data accessible to

anyone with the technical skill to run a Shodan search.

An open MongoDB containing the location histories and personal details of 5 million

users is not a sophisticated attack; it is a failure to implement the most

elementary security controls. The Dalil breach predated the PDPL, but it exemplifies

the data protection failures the law was designed to prevent. Any application

collecting personal data in Saudi Arabia today must treat database security as a

non-negotiable minimum, not an afterthought.

RELATED ANALYSIS

Cisco Systems: ShinyHunters Claim 3M Salesforce Records, 300+ GitHub Repos, and AWS Data in Triple-Vector Extortion
Mar 31, 2026 · 3M+ records claimed · 300+ repos · April 3 deadline
Oracle's Dual Breach: 6M Cloud SSO Records Stolen, 80 Hospitals Compromised - and a Denial That Collapsed Under Evidence
Mar 21, 2025 · 6M records · 140K tenants · 80 hospitals
TriZetto/Cognizant: 3.4M Patient Records Stolen in 11-Month Healthcare Supply Chain Breach
Feb 6, 2026 · 3.4M patients · 11-month dwell · ~24 lawsuits
Infinite Campus: ShinyHunters Breach K-12 Platform Serving 11M Students via 10-Minute Vishing Attack
Mar 18, 2026 · 11M students · 3,200+ districts · 46 states
Crunchyroll: 6.8M Users Exposed After Infostealer Malware Compromises TELUS Support Agent's Okta Credentials
Mar 12, 2026 · 6.8M users · 100GB stolen · $5M ransom
MORE DATA BREACHES →