×
Have questions or ready to talk to a Vonage expert?
Robot Chat Icon
Device Type: 
Skip to Main Content Skip to Main Content

AIT vs. SMS Pumping vs. Toll Fraud: What is the Difference and How To Fight Each

This article was published on November 19, 2026

SMS pumping is the SMS-only branch of artificially inflated traffic, which is itself powered by the toll-fraud revenue-share machinery known as IRSF. The three overlap in mechanics but demand different defense levers, and knowing which is which decides where your budget goes.

 

Digital globe with orange routes linking regions worldwide.
Adam Weir headshot
By Adam Weir Product Expert, Fraud Solutions

Key takeaways 

  • SMS pumping is a subset of artificially inflated traffic (AIT). AIT is the parent category across SMS and voice. Toll fraud (IRSF) is the revenue-sharing engine that pays the attacker.
  • The mechanics overlap. Vendors blur the labels, but each family has a distinct defense lever, and mismatching lever to attack wastes fraud budget.
  • Telecom fraud reached $41B in 2025 , and up to 60% of OTP requests can be fraudulent. Exposure is measured in dollars per hour, not per quarter.
  • Pre-send scoring, during-send burst controls, and prefix and IRSF rules each stop a different piece. Fraud Defender ties them into one decision.
  • In one anonymized hyperscale deployment, unified defense hit a 90% detection rate, blocked 4.6M messages over six months, and documented $186K in savings.

Why the labels matter

Heads of Fraud, CISOs, and the platform engineers who own messaging hear these three terms used as if they were synonyms. They are not. AIT, SMS pumping, and toll fraud describe overlapping plumbing, yet the control that stops one may do nothing against another.

The stakes are not abstract. Data from the CFCA put global telecom fraud losses at $41B in 2025 . Meanwhile AB Handshake research indicates that up to 60% of OTP requests can be fraudulent during an active traffic-pumping campaign. For a hyperscale sender, that translates into $10K or more disappearing inside a single six-hour window before anyone reads the alert.

Getting the taxonomy right does three things for a fraud team. It lets you allocate budget to the lever that actually moves your loss curve. It lets you vet vendors against the specific attack you face instead of a generic "fraud" claim. And it gives you the language to explain the threat to a CFO or board sponsor who signs the check.

AIT (artificially inflated traffic), defined

Artificially inflated traffic is the umbrella pattern. Bots trigger authentication and verification of events. Events like one-time passcodes, sign-up confirmations, or password resets. This inflates your messaging spend while a fraudster collects a cut of the termination fees. Because it rides your own legitimate flows, AIT looks like organic growth until the invoice arrives.

AIT is a form of communications fraud that targets any budget attached to programmatic messaging. It spans both SMS and voice, and it concentrates on the flows that any new or anonymous user can reach without authentication. Including the signup form, the "resend code" button, and the standard phone-verification step. That is the same surface where OTP abuse lives, which is why teams see AIT and OTP abuse spike together.

The tell is a divergence between traffic and outcomes. Message volume, and the bill, climb steeply while completed sign-ups, activations, or logins stay flat. If your OTP send count doubles but your verified-account count does not, you are very likely funding artificially inflated traffic.

Fraudsters sent an estimated 12.94 billion artificially inflated traffic messages in 2024, costing brands $1.06 billion .

SMS pumping, defined

SMS pumping is the SMS-specific subset of AIT. Where AIT can also exploit voice, pumping fraud in this narrower sense uses SMS-only flows. Bots harvest revenue by forcing your platform to send paid text messages to number ranges the attacker controls or profits from. Every pumped message is real, delivered, and billable, which is exactly why SMS pumping is hard to spot with naive volume alarms.

The economics are brutal for high-volume senders. Because SMS pumping abuses verification and onboarding endpoints, it clusters in the same free-trial, account-creation, and passcode flows that drive your acquisition funnel. Attackers favor destinations and prefixes where termination rates are high. Leading to loss per pumped messages often far above your average route cost.

SMS pumping and broader SMS fraud are frequently reported as one number in industry surveys, which understates how targeted the SMS variant is. Treating SMS pumping as its own line item, with its own burst thresholds and geographic permissions, is the first step to controlling it.

Toll fraud (IRSF), defined

Toll fraud, in the communications context, is International Revenue Share Fraud (IRSF). It is not a separate "third attack" so much as the underlying payout mechanism that makes both AIT and SMS pumping profitable. A fraudster acquires access to premium-rate or high-cost international number ranges, arranges a revenue-share deal on the traffic terminating there, then drives your platform to route messages or calls to those ranges. The termination fees flow back to the fraudster.

That is why toll fraud spans SMS and voice alike. Any traffic your platform sends to an attacker-linked premium-rate destination can be monetized. IRSF is the engine while AIT and SMS pumping are two of the vehicles bolted onto it. People may ask how toll fraud relates to AIT. The honest answer is that AIT is usually a delivery method for IRSF, and SMS pumping is the SMS-flavored version of that method.

Because IRSF is anchored to specific number ranges and prefixes, it is also the family most amenable to hard, deterministic blocking. Known fraudulent ranges can be denied outright, and high-risk destinations can be gated before a single message leaves your gateway.

Side-by-side: The communications fraud comparison

The fastest way to keep the families straight is to compare them by channel, target, and detection method.

Aspect

AIT (artificially inflated traffic)

SMS pumping

Toll fraud / IRSF 

Channel 

SMS and voice 

SMS only 

SMS and voice 

What it targets 

Total enterprise messaging budget 

Enterprise SMS budget specifically

Any traffic routed to premium-rate or high-cost ranges 

Primary flows abused 

Signup, verification, OTP, resets 

Signup, verification, OTP (SMS) 

Any flow reaching a revenue-share destination 

Payout mechanism

IRSF revenue sharing 

IRSF revenue sharing 

IRSF revenue sharing (the mechanism itself) 

Core detection 

IRSF databases plus sequential dialing detection 

IRSF databases, sequential dialing, plus burst controls 

Prefix rules plus IRSF databases 

Relationship 

Parent category 

Subset of AIT 

Underlying engine for both 

The single most useful insight in the table. Notice the detection rows overlap heavily. IRSF datasets and sequential-dialing intelligence recur across all three, which is precisely why a unified signal graph beats three-point tools bought from three suppliers.

Defense lever by attack type: Matching communications fraud defense to the threat

Each family responds to a different primary lever. Buy the wrong one and you pay for coverage you cannot use.

Stopping AIT requires pairing pre-send risk scoring with continuous during-send controls. The strongest position is to check every message against prefix rules, IRSF matching, and sequential-dialing intelligence before it is sent, so risky traffic can be held or blocked before it is billed . During-send controls in the Advanced tier catch what shifts mid-campaign.

Defeating SMS pumping requires pairing active burst protection plus geo permissions. Because pumping fraud manifests as sudden SMS volume toward specific destinations, the counter is SMS burst thresholds combined with geographic permissions, allowing or denying traffic by country and range, in Fraud Defender Advanced. This is where an SMS-specific line item pays off.

Stopping Toll fraud and IRSF comes down to strict prefix rules plus IRSF databases. The deterministic layer. Auto-defined prefix rules block high-risk countries, and customer-defined rules override them to allow or deny specific ranges, from a single number up to a whole country. Behind the prefix layer, the AIT Intelligence Database checks third-party IRSF datasets refreshed as often as every five minutes, returning a "major match" (close match to a confirmed fraudulent range, last three digits) or a "minor match" (a wider fraudulent range, last four digits).

Underneath all three sits Vonage's own Sequential Dialling Detection, continuously updated with entries pruned after 180 days, scoring traffic against pumping intelligence by attack size, age, and recurrence.

The cleanest way to shrink the attack surface for artificially inflated traffic and SMS pumping is to stop leaning on interceptable passcodes in the first place. Shifting away from legacy OTP removes the underlying messaging channel that fraudsters exploit to generate fake delivery fees. Pairing your messaging defense with Silent Authentication, part of the Verify API, removes the OTP that bots harvest. For account takeover risk specifically, add SIM Swap Detection. Silent auth alone removes the interceptable code, but full SIM-swap protection needs the detection signal alongside it.

The unified Fraud Defender approach: One signal graph across every attack type

The reason point tools leak is structural. No single CPaaS supplier sees all of your traffic. Run two or more messaging vendors and fraud simply migrates into the gaps between their siloed fraud products. Each supplier scores only its own slice. Nobody holds the full picture.

Fraud Defender inside Network-Powered Solutions (Network API) closes that gap. It is the unified communications-fraud defense layer, one vendor, one signal graph. For traffic riding Vonage, the Standard, Advanced, and Premium tiers layer prefix rules, IRSF matching, burst protection, and sequential-dialing intelligence together. 

AIT cost brands an estimated $1.06 billion in 2024, driven by roughly 12.94 billion fraudulent messages, and Mobilesquared puts it at 3.23% of global A2P SMS spend Fraud Defender Advanced customers have already benefited from that defense, with over $3.8 million in documented savings to date and blocked-traffic costs cut by up to 90%. 

Fraud Defender Advanced has already saved customers more than $3.8 million by blocking fraudulent traffic before it's billed, cutting costs by up to 90% on blocked volume .

How to assess your exposure

You can gauge your risk before you talk to any vendor. Three checks surface the vast majority of active AIT and SMS pumping campaigns.

Audit your SMS spend trend. Plot monthly messaging cost against active users or transactions. A cost curve that outpaces user growth is the signature of pumping fraud draining budget through legitimate-looking sends.

Compare OTP volume to successful sign-up ratio. Divide completed verifications by OTP requests. A healthy funnel converts most requests. If a large share of codes are requested and never used, you are almost certainly funding OTP abuse and artificially inflated traffic.

Identify markets disproportionate to your customer base. Rank message volume by destination country and range, then compare against where your real customers live. Traffic concentrating in markets where you have no meaningful user base points straight at IRSF revenue-share destinations.

Communications fraud comparison FAQs

Select to expand or collapse this FAQ answer.

Artificially inflated traffic is bot-driven abuse that forces your platform to send high volumes of paid messages or calls, usually through verification and OTP flows. A fraudster collects a share of the termination fees. It is the parent category above SMS pumping and spans both SMS and voice.

Select to expand or collapse this FAQ answer.

SMS pumping is the SMS-only subset of AIT. Bots trigger paid text messages to number ranges the attacker profits from, harvesting revenue while your SMS bill climbs. Every pumped message is real and billable, which is what makes it hard to catch with simple volume alarms.

Select to expand or collapse this FAQ answer.

Toll fraud in messaging is the abuse of premium-rate or high-cost international destinations to generate revenue for the attacker. It is the payout layer that both AIT and SMS pumping rely on, and it spans SMS and voice.

Select to expand or collapse this FAQ answer.

IRSF, International Revenue Share Fraud, is the revenue-sharing arrangement behind toll fraud. A fraudster secures a cut of the fees on traffic terminating at specific number ranges, then drives your platform to route messages or calls there. IRSF is the engine while AIT and SMS pumping are two vehicles that use it.

Select to expand or collapse this FAQ answer.

No, though vendors often use the terms interchangeably. AIT is the broader category across channels and SMS pumping is a subset. IRSF is the underlying money mechanism both exploit. The mechanics overlap, but the defense levers differ. 

Select to expand or collapse this FAQ answer.

You can neutralize artificially inflated traffic by scoring risk pre-send with prefix rules, IRSF matching, and sequential dialing detection, and deploying  tight, during-send validation controls. Countering SMS pumping requires an entirely different setup, pairing active burst protection with strict geographic permissions. Finally, you can completely block toll fraud and IRSF schemes by enforcing granular prefix rules backed by live, up-to-date threat databases. A unified layer with one signal graph ties all three together and reduces OTP abuse feeding them. 

Select to expand or collapse this FAQ answer.

It is a continuously updated intelligence layer that scores traffic against known traffic-pumping behavior, weighing attack size, age, and recurrence, with entries pruned after 180 days. It runs beneath prefix and IRSF checks to catch pumping patterns those rules alone would miss. 

Deskphone with Vonage logo
Outside the US: Local Numbers