VoIP Toll Fraud and SIP Security: Attacks, Warning Signs and Controls
Toll fraud is the abuse of a telephone system to place calls — usually to expensive international or premium-rate numbers — at the victim's expense. On IP telephony it almost always starts with an attack on SIP, the signalling protocol of VoIP, and it can cost thousands of euros in a single night.
Key takeaways
- Internet-exposed SIP services are scanned continuously; weak extension passwords are found within hours.
- International Revenue Share Fraud (IRSF) turns stolen call capacity into money for the attacker.
- Most fraud happens at night and at weekends, when nobody watches the call records.
- Layered controls — strong credentials, rate limiting, geo and destination restrictions, call limits and real-time monitoring — stop the vast majority of attacks.
How an attack unfolds
- Scanning — automated tools sweep the internet for SIP services on UDP/TCP 5060 and 5061 and identify the PBX type.
- Extension enumeration and brute force — the attacker discovers valid extensions and tries passwords until one works.
- Registration or trunk abuse — the attacker registers as the extension, or abuses an open route or a misconfigured trunk.
- Monetisation — calls are pumped to premium or high-cost international numbers that share revenue with the attacker (IRSF), often many in parallel.
Main threats to SIP and VoIP
| Threat | Description | Impact |
|---|---|---|
| Toll fraud / IRSF | Calls to premium or high-cost destinations through a compromised account or route | Direct financial loss |
| SIP brute force | Password guessing on REGISTER and INVITE | Account takeover |
| Registration hijacking | Replacing a legitimate device registration to receive its calls | Interception, fraud |
| Call and message flooding | Floods of INVITE or other SIP requests | Denial of service |
| Malformed SIP messages | Crafted packets targeting parser vulnerabilities | Crash, potential compromise |
| Caller ID spoofing | Falsified calling number used for fraud and vishing | Social engineering, reputational damage |
| Eavesdropping | Unencrypted signalling and media | Confidentiality loss |
Warning signs in call records
- Sudden spikes of outbound calls, especially outside business hours.
- Calls to unusual international destinations or premium-rate ranges.
- Many simultaneous calls from a single extension or trunk.
- Bursts of failed registrations or authentication errors from the same address range.
- Very short repeated calls (testing) followed by long calls.
Controls that work
Access
- Long, random SIP passwords, different from extension numbers; no default credentials.
- Do not expose SIP to the whole internet if not needed: VPN, allow-lists or a session border controller.
- Geo-filtering and ASN filtering of SIP traffic to the countries and networks you actually use.
- Rate limiting and automatic banning of brute-force sources (for example the Kamailio
pikemodule and fail2ban).
Dialling rules
- International and premium-rate destinations disabled by default, enabled per user only when needed.
- Per-extension and per-trunk limits on simultaneous calls and on spend per hour or day.
- Stricter rules outside business hours; secure PINs on voicemail and remote-access (DISA) features.
Protection and monitoring
- TLS for signalling and SRTP for media.
- Patching of the PBX, SBC and phones; removal of unused modules.
- Real-time analysis of call records with alerting and automatic blocking of suspicious calls.
- Logs sent to the SIEM so that VoIP incidents are handled by the SOC.
Caller ID authentication
Operators increasingly authenticate calling numbers using STIR/SHAKEN-type mechanisms. France introduced mandatory number authentication for operators in 2024. Enterprises still need their own controls, because their PBX and SIP accounts remain the entry point for fraud.
Audit your exposure
A VoIP security audit checks what is reachable from the internet, the strength of credentials, dialling rules, encryption, firmware versions and monitoring. Only test systems you own or are authorised to assess.
Frequently asked questions
What is VoIP toll fraud?
The abuse of a phone system to make calls, usually to premium or expensive international numbers, at the owner's expense. On IP telephony it typically starts with a compromised SIP account or a misconfigured route.
What is IRSF?
International Revenue Share Fraud: attackers generate calls to high-cost international numbers that share part of the call revenue with them.
How do I protect an Asterisk or IP-PBX against SIP brute force?
Use strong random passwords, avoid exposing SIP publicly or restrict it by IP, country and network, rate-limit and ban abusive sources, and monitor failed registrations.
When does toll fraud usually happen?
Mostly at night, at weekends and during holidays, when nobody watches call activity — which is why real-time automatic detection and blocking matter.
Related guides
- AI SOC Automation — How AI agents assist SOC analysts, with an open-source stack.
- AI Observability — What to monitor in LLM apps and agents, with OpenTelemetry.
- Shadow AI — Discover and govern unapproved AI, agents and MCP servers.