FTC Consumer Sentinel Network: 16 Million Fraud Reports Hiding in Plain Sight

Every year, Americans lose tens of billions of dollars to fraud. The Federal Trade Commission's Consumer Sentinel Network is the most comprehensive public record of those losses — more than 8 million reports annually from the FTC itself and dozens of partner organizations including the Better Business Bureau, state attorneys general, the CFPB, and the Social Security Administration. The cumulative dataset now exceeds 16 million records. Almost no one outside law enforcement analyzes it at full scale.

This article covers what Consumer Sentinel contains, how to obtain the data, what the aggregate findings reveal about fraud patterns in the United States, how journalists and researchers use it, a Python analysis walkthrough, and how the network directly informs FTC enforcement priorities and rulemaking.

What Consumer Sentinel is and who feeds it

Consumer Sentinel began in 1997 as an internal FTC database for tracking telemarketing fraud complaints. Over the following two decades it grew into a federated network: the FTC accepts complaint submissions directly through ReportFraud.ftc.gov and its predecessor consumer.ftc.gov, and partner organizations contribute their own complaint files under data-sharing agreements. As of 2025 the network has over 35 partner contributors.

Major data contributors beyond the FTC itself include the Internet Crime Complaint Center (IC3) operated by the FBI, which submits cybercrime complaints; the CFPB, which submits financial product complaints; the Social Security Administration Office of the Inspector General, which submits Social Security impostor complaints; the BBB Scam Tracker; state attorneys general from roughly 40 states; the Canadian Anti-Fraud Centre; and the National Do Not Call Registry for robocall complaints. Each partner contributes under its own complaint taxonomy, which the FTC maps to a standardized category hierarchy before loading into Sentinel.

The practical consequence of this architecture is that Consumer Sentinel captures fraud that consumers never reported to the FTC directly. A consumer who filed a BBB Scam Tracker report on a fake online retailer, a CFPB complaint against a debt collection firm, and a state AG complaint against a contractor may be represented in Sentinel three times, though the FTC does deduplicate across some partner feeds. The network's breadth is its core analytical advantage over any single-agency complaint database.

Dataset structure: what each record contains

Each Consumer Sentinel record represents a single consumer complaint and contains the following core fields:

Reported amount is the most analytically important field and the most distorted. It reflects what the consumer believed they lost at the time of filing, which may differ from actual financial loss due to chargebacks, partial recovery, or simple inaccuracy. For categories like investment fraud and romance scams, where losses accumulate over months, consumers who file early may substantially understate total losses. The FTC's published aggregate dollar figures should be read as lower bounds.

How to access the data

Consumer Sentinel data is available at two levels of granularity, each through a different access pathway.

Public aggregate reports

The FTC publishes annual Consumer Sentinel Network Data Books at ftc.gov/enforcement/consumer-sentinel-network. These PDF reports contain national-level and state-level breakdowns of complaint counts and reported dollar losses by category, payment method, age group, and contact method. The Data Books are the fastest entry point for journalists and policy analysts who need aggregate numbers without building a data pipeline. The FTC publishes the underlying data tables as Excel files alongside each PDF, making the aggregate statistics machine-readable.

The FTC also maintains an interactive data visualization at consumer.ftc.gov/consumer-sentinel-network-data-book, which allows users to filter by state, year, and complaint category. The visualization tool is useful for quick state-by-state comparisons but does not expose all category combinations and does not support bulk export.

Bulk record-level access

Record-level Sentinel data is not publicly downloadable. Law enforcement agencies can apply for direct network access through the FTC's law enforcement portal. Journalists and researchers without law enforcement affiliation have two practical paths: filing a FOIA request for de-identified complaint records, or working with a law enforcement agency partner that has Sentinel access.

The FTC does release de-identified flat files for specific research purposes under its data sharing program. These files omit direct identifiers (name, address, email, phone number) but retain all analytical fields including reported amount, payment method, age range, state, contact method, and complaint category. The FTC has released such files to academic researchers studying specific fraud categories under data use agreements that restrict publication of individual records and require results review before public release.

For journalists, the most productive access path is the annual Data Book Excel files combined with FOIA requests targeted at specific complaint subcategories or specific time periods. The FTC processes Sentinel FOIA requests under the law enforcement exemption (Exemption 7) with some regularity, releasing de-identified aggregate files that go beyond what the Data Books publish.

Top fraud categories: what Americans report most

Three complaint categories have dominated Consumer Sentinel for the past five years and together account for the majority of reported dollar losses.

Imposter scams are the single largest fraud category by both complaint count and reported losses. Fraudsters impersonate government agencies (the IRS, Social Security Administration, Medicare, the FTC itself), technology companies (Microsoft, Apple, Amazon), financial institutions, and increasingly, romantic partners or family members in distress. The Social Security impostor sub-category alone generates several hundred thousand reports annually. Government impostor calls typically demand payment via gift card or wire transfer, threatening arrest or benefit termination. In 2023, imposter scams accounted for over 850,000 reports and more than $2.7 billion in reported losses.

Online shopping and negative option fraudis the second-largest category by complaint volume. This category encompasses fake retailers that take payment and ship nothing, counterfeit goods, subscription services with undisclosed auto-renewal terms, and “free trial” offers that convert to recurring charges. The explosion of social media advertising has substantially amplified this category: the FTC's data shows that ads on Facebook and Instagram are the leading contact method for online shopping fraud, accounting for roughly 45% of reported social media fraud losses across all categories in recent data.

Identity theft is tracked separately from fraud complaints and is the largest single component of Consumer Sentinel by raw record count. The FTC's IdentityTheft.gov processes identity theft reports and feeds them into Sentinel. Government benefits identity theft — where a fraudster uses a victim's Social Security number to file for unemployment insurance, stimulus payments, or pandemic-era relief programs — spiked dramatically in 2020 and 2021 before declining as pandemic programs wound down. Credit card fraud and loan or lease fraud are the other dominant sub-categories within identity theft.

Payment methods: where the money goes

Payment method is one of the most policy-significant fields in Consumer Sentinel because payment rails have different chargeback rights, traceability, and recovery rates. The FTC tracks payment method data systematically because it directly informs which payment processors and financial institutions receive enforcement scrutiny.

Bank wire transfer produces the highest median loss per report of any payment method — typically over $10,000 per incident — because fraudsters who can persuade a victim to wire money are typically running high-value investment fraud, business email compromise, or romance scams. Wire transfers are irreversible once settled and are nearly impossible to recover. Investment fraud, which is heavily concentrated among victims who wire funds to offshore accounts, accounts for a disproportionate share of total reported wire transfer losses.

Cryptocurrency has become the second-highest median-loss payment method and the fastest-growing. The FTC reported that consumers lost over $1.4 billion to cryptocurrency fraud in 2023, up from $130 million in 2020. The dominant sub-category is investment fraud using fake trading platforms: victims are shown fabricated portfolio balances, asked to “invest more to unlock withdrawals,” and ultimately left with nothing when the platform disappears. These schemes, sometimes called “pig butchering” (from the Chinese “sha zhu pan”), frequently originate in Southeast Asian fraud compounds and are particularly prevalent in Sentinel records with initial contact via social media or dating applications.

Gift cards produce the highest complaint count among non-card payment methods. Imposter scams — IRS, Social Security, tech support — disproportionately demand gift card payment because the cards are available at retail locations familiar to older consumers, are difficult to trace once redeemed, and have no chargeback mechanism. The median gift card fraud loss is lower than wire transfer or cryptocurrency but the volume is substantially higher, making gift cards a significant contributor to aggregate reported losses. The FTC has pressed major retailers including Target, CVS, and Walgreens to add point-of-sale friction when customers purchase large numbers of gift cards, which some stores have implemented through employee training programs.

Age demographics: who reports more, who loses more

The relationship between age and fraud victimization in Consumer Sentinel data is counterintuitive and frequently misreported. The data shows two distinct patterns that are often conflated.

Younger consumers in the 20–39 age range report fraud at higher rates per capita than older consumers. This reflects both higher exposure (more online shopping, more social media use, more engagement with digital financial products) and higher willingness to file complaints. The FTC's analysis of 2023 data found that people in their 30s reported fraud more often than any other age decade.

Older consumers, particularly those 70 and above, report less fraud per capita but lose substantially more money when they are defrauded. The median reported loss for consumers over 70 is consistently two to four times higher than for consumers under 40. Investment fraud and romance scams — both of which involve sustained manipulation over weeks or months and produce very large individual losses — are disproportionately reported by consumers over 60. The FTC's analysis attributes the higher median loss among older consumers to several factors: greater accumulated savings, higher willingness to trust unfamiliar contacts, and less familiarity with the warning signs of cryptocurrency and investment platforms.

The implication for the age field in Consumer Sentinel analysis is that complaint count and reported dollar loss should not be analyzed together without controlling for the age-specific fraud type distribution. A state with a high proportion of older residents will tend to show higher median losses per report but lower report rates per capita.

State-level patterns: per-capita complaint rates

The FTC publishes per-capita complaint rankings by state in each annual Data Book. The rankings are normalized per 100,000 population using U.S. Census Bureau estimates and cover both total complaints and identity theft separately. The per-capita normalization is important because raw complaint counts are heavily correlated with state population.

Florida and Georgia consistently rank among the top five states for per-capita fraud complaints. Both states have large retiree populations (higher median losses), major international airports that serve as entry points for fraud operations, and historically lower state-level consumer protection enforcement capacity relative to their population. Nevada, Delaware, and Maryland round out the frequent top performers, though their mechanisms differ: Nevada has a high transient population and heavy online gambling activity; Delaware's high business incorporation rate creates entity infrastructure that fraud operations exploit; Maryland's high federal employment concentration makes it a target for government impostor schemes.

Identity theft per-capita rankings show a different geographic pattern, with more concentration in states that had large pandemic-era unemployment insurance programs and weaker identity verification requirements. States with more sophisticated ID proofing requirements for public benefits show systematically lower identity theft report rates in Sentinel, a finding the FTC has cited in policy guidance on benefits fraud prevention.

Python analysis: loading and grouping by category and payment method

The following example demonstrates loading a de-identified Consumer Sentinel flat file and computing the key cross-tabulations that the FTC publishes in its Data Books. This assumes access to a record-level extract through a data use agreement or FOIA release.

import pandas as pd

# Load de-identified Consumer Sentinel flat file
# Typical fields: report_date, category, sub_category, reported_amount,
#                 payment_method, contact_method, age_range, state, source
df = pd.read_csv("sentinel_deidentified_2024.csv", low_memory=False,
                 parse_dates=["report_date"])

# --- 1. Complaint count and reported loss by category ---
cat_summary = (
    df.groupby("category")
    .agg(
        complaints=("category", "count"),
        total_loss=("reported_amount", "sum"),
        median_loss=("reported_amount", "median"),
        pct_with_loss=("reported_amount", lambda s: (s.notna() & (s > 0)).mean() * 100)
    )
    .sort_values("complaints", ascending=False)
    .reset_index()
)
print("Top fraud categories by complaint volume:")
print(cat_summary.head(10).to_string(index=False))

# --- 2. Reported loss by payment method ---
pay_summary = (
    df[df["reported_amount"].notna() & (df["reported_amount"] > 0)]
    .groupby("payment_method")
    .agg(
        complaints=("payment_method", "count"),
        total_loss=("reported_amount", "sum"),
        median_loss=("reported_amount", "median")
    )
    .sort_values("total_loss", ascending=False)
    .reset_index()
)
print("\nLosses by payment method (sorted by total reported loss):")
print(pay_summary.to_string(index=False))

# --- 3. Median loss by age range ---
age_order = ["20-29", "30-39", "40-49", "50-59", "60-69", "70-79", "80+"]
age_summary = (
    df[df["age_range"].isin(age_order) & df["reported_amount"].notna() & (df["reported_amount"] > 0)]
    .groupby("age_range")
    .agg(
        complaints=("age_range", "count"),
        median_loss=("reported_amount", "median"),
        mean_loss=("reported_amount", "mean")
    )
    .reindex(age_order)
    .reset_index()
)
print("\nMedian reported loss by age range:")
print(age_summary.to_string(index=False))

# --- 4. Category x payment method cross-tab (top 5 categories) ---
top_cats = cat_summary["category"].head(5).tolist()
cross = (
    df[df["category"].isin(top_cats) & df["payment_method"].notna()]
    .groupby(["category", "payment_method"])
    .agg(complaints=("category", "count"), total_loss=("reported_amount", "sum"))
    .reset_index()
    .sort_values(["category", "total_loss"], ascending=[True, False])
)
print("\nCategory x payment method breakdown (top 5 categories):")
print(cross.to_string(index=False))

The category-by-payment-method cross-tabulation is where Consumer Sentinel analysis diverges most clearly from the FTC's own published summaries. The Data Books report category and payment method in separate tables but do not publish the interaction — which payment methods are most associated with which fraud categories. That cross-tabulation is analytically important because payment method is the strongest predictor of recovery probability and is the basis for the FTC's outreach to specific payment processors.

How journalists and researchers use Consumer Sentinel

The annual Data Books have generated substantial investigative journalism, much of it focused on the role of technology platforms as fraud infrastructure. The FTC's finding that social media is the contact channel for the largest aggregate reported losses among all contact methods prompted congressional hearings and was cited in litigation against Meta by state attorneys general. The FTC itself used Social Sentinel data on Instagram-initiated investment fraud to support its 2022 enforcement action against a cryptocurrency promoter network.

Academic researchers have used Sentinel data primarily to study the demographic dimensions of fraud vulnerability. A series of papers using de-identified Sentinel extracts has examined how the probability of being defrauded interacts with financial literacy, cognitive decline, social isolation, and prior victimization. The finding that prior fraud victims are significantly more likely to appear multiple times in Sentinel than random consumers — a pattern sometimes called “sucker list resale” — has been replicated across multiple data extracts and is consistent with the observed market for fraud lead lists in law enforcement investigations.

State attorneys general use Sentinel data operationally to identify complaint clusters that indicate an active fraud operation. If a specific company, phone number, or website suddenly accumulates complaints from multiple states, the FTC surfaces the pattern to relevant state AGs through the Sentinel law enforcement portal. Several major multi-state enforcement actions — including actions against student loan debt relief firms and extended vehicle warranty telemarketers — originated in Sentinel complaint clusters before the FTC or states had enough individual complaints to pursue enforcement unilaterally.

Policy significance: how Sentinel shapes FTC enforcement priorities

Consumer Sentinel is not merely a historical record. The FTC uses it as a forward-looking tool for identifying emerging fraud patterns and setting enforcement priorities. The mechanism is systematic: FTC staff analysts review complaint trends on a rolling basis, flag categories with rapid growth or unusual loss concentration, and trigger preliminary investigative inquiries when patterns meet internal thresholds.

The cryptocurrency investment fraud surge is the clearest recent example of this pipeline. Sentinel showed a roughly tenfold increase in cryptocurrency-related complaint volume between 2020 and 2022, with a disproportionate concentration in reports describing fake trading platforms and romance-initiated contacts. This data drove the FTC's 2022 Consumer Alert on cryptocurrency investment scams, its coordinated referrals to the Department of Justice and FBI, and its enforcement actions under Section 13(b) of the FTC Act against operators of fraudulent trading platforms.

Sentinel data also informs the FTC's rulemaking function. The 2023 Non-Compete Clause Rule and the 2024 proposed rule on junk fees both cited Consumer Sentinel complaint data on deceptive pricing and undisclosed terms as empirical support for the need for regulatory action. The FTC's Negative Option Rule update — governing subscription cancellation requirements — drew heavily on the volume and consistency of negative option complaints in Sentinel to establish that the existing rule was inadequate.

The network's federated structure amplifies its policy reach. Because state AGs contribute to and access Sentinel, the FTC's enforcement priorities can be replicated at the state level without requiring duplicative complaint collection. A pattern the FTC identifies nationally can be investigated locally by state AGs using their own state UDAP (Unfair and Deceptive Acts and Practices) authority. This coordination explains why many major consumer fraud enforcement actions are announced as joint federal-state actions even when the investigative lead is the FTC.

For researchers and journalists, the implication is that Consumer Sentinel is not a static archive. It is the input to an ongoing enforcement process. Understanding which categories are growing in complaint volume and which payment methods are generating the largest losses — before the FTC's annual Data Book publication — provides advance signal on where FTC enforcement attention is likely to focus in the following 12 to 18 months. The FTC's enforcement lag between complaint pattern identification and formal action is typically 18 to 36 months, which means the 2024 and 2025 complaint data will largely determine the enforcement docket of 2026 and 2027.