2026 Chargeback Fraud Trends: What Merchants Need to Prepare For

Trend 1: Agentic Commerce Creates Undetectable Fraud

AI agents completing purchases — without human review of each transaction — are generating chargebacks that look legitimate because they are legitimate at the authentication layer. Adobe Analytics recorded a 4,700% year-over-year jump in AI-driven traffic to US retail sites in early 2026, meaning a material share of e-commerce transactions are now agent-initiated.

The fraud problem: traditional fraud signals break entirely for agent transactions.

• Device fingerprinting fails — an AI agent shopping from a cloud server looks like a data center IP, not a consumer device
• Behavioral biometrics fail — there is no human mouse movement, typing cadence, or hesitation pattern
• IP geolocation fails — cloud servers can be anywhere, and proxy/VPN detection rules misfire

Card networks have no liability framework for agent-authorized transactions yet. When a consumer disputes a purchase their AI agent made, there is no clear chargeback reason code — and no clear evidence framework to fight it. Merchants taking agent-initiated orders should add explicit consumer confirmation steps in 2026.

Trend 2: First-Party Fraud Surges to 36% of All Disputes

First-party fraud — customers disputing legitimate purchases — has grown from 15% of fraud in 2023 to 36% in 2026. Economic pressure is the primary driver: consumers under financial stress increasingly treat chargebacks as an accessible refund mechanism, particularly for digital goods and subscriptions where perceived harm is low.

Visa CE 3.0 (Compelling Evidence 3.0) is the primary tool to fight first-party fraud. By demonstrating two prior non-disputed transactions from the same device, IP address, and account, merchants shift liability back to the issuing bank. Building CE 3.0 documentation as standard practice is no longer optional — it is the primary defense against this category.

Trend 3: BNPL Chargebacks Growing Alongside BNPL Adoption

Buy Now Pay Later adoption continues to grow across retail and travel segments, and dispute volumes are growing proportionally. A new category of chargeback is emerging: consumers disputing BNPL installments they decide to stop paying — not because of fraud, but because they no longer want the item or regret the purchase.

The structural challenge: liability in BNPL chargebacks is still evolving. When a consumer disputes a BNPL installment:

• Some BNPL providers absorb the chargeback and pursue the merchant separately
• Others pass the chargeback directly to the merchant via the card network
• The processing rules vary by BNPL provider, card network, and acquiring bank

Merchants selling via BNPL need separate dispute response strategies for each BNPL provider — the evidence requirements and timelines differ from standard card disputes.

Trend 4: Visa VAMP Threshold Drops to 0.9%

In April 2026, Visa launched VAMP (Visa Acquirer Monitoring Programme), replacing the older VDMP and VFMP programmes. The new thresholds are significantly tighter than what merchants managed against previously.

This is approximately 40% tighter than previous thresholds. Merchants operating at 1.2–1.5% under the old VDMP rules are now in Early Warning territory. The margin for error is substantially smaller in 2026.

Use our VAMP calculator to check your current exposure before enforcement notices arrive.

Trend 5: Mastercard SMMP Scam Merchant Rules (July 2026)

Mastercard's SMMP (Scam Merchant Monitoring Programme) adds a new dimension to chargeback management: merchants can now be flagged not just for high chargeback ratios, but for being used in scam ecosystems — even without direct fault.

Active enforcement begins July 2026. SMMP targets merchants with high rates of scam-related disputes, including:

• Triangulation fraud: Merchants whose products appear in orders placed with stolen payment credentials
• Reshipping schemes: Merchants unknowingly shipping to reshipping mule addresses
• Third-party scam facilitation: Any pattern where a disproportionate share of disputes are scam-coded

Use our SMMP calculator to assess your exposure. If your Mastercard scam dispute rate is elevated, proactive remediation before July 2026 is significantly better than responding to an enforcement notice.

Trend 6: SaaS Chargebacks Up 83% Year-Over-Year

Subscription and SaaS chargebacks grew 83% year-over-year — the fastest growing dispute category in 2026. Three primary drivers:

Mastercard's 2026 subscription billing rules require explicit pre-charge notification for all recurring billing merchants. Non-compliance means automatic chargeback loss — no representment pathway — for disputes where notification was not sent.

Trend 7: AI-Powered Fraud Detection Is Also Improving

The same AI capabilities enabling new fraud patterns are also improving fraud detection. Merchants who adopt AI-powered fraud tools in 2026 have a meaningful advantage over those relying on static rule sets.

Key improvements in AI fraud detection:

• Machine learning pattern recognition: Models trained on cross-merchant fraud data identify patterns that single-merchant rule sets cannot detect
• Behavioral biometrics at scale: New models distinguish AI agents from human shoppers through session-level signal patterns even without traditional device fingerprints
• Adaptive rules: Fraud tools that update automatically as new attack patterns emerge, rather than requiring manual rule updates
• Network-level signals: Shared fraud intelligence across merchant networks means one merchant's fraud incident becomes another's prevention signal

The practical implication: merchants using AI fraud tools will recover faster from new attack patterns than those using static rule sets — because the tools adapt, rather than requiring manual intervention each time a new vector emerges.

How to Prepare for 2026 Chargeback Trends

A practical preparation checklist based on the seven trends above:

1. Implement CE 3.0 documentation for all digital transactions. Capture and retain device fingerprint, IP address, account creation date, and session logs for every order — not just disputed ones. CE 3.0 requires prior transaction records, which means starting now.
2. Add a human confirmation step for AI/agent-initiated purchases. If your platform accepts agent-initiated orders (shopping APIs, autonomous checkout), add an explicit human verification step or clear authorization confirmation before fulfilling high-value orders.
3. Check your current Visa VAMP ratio. Use the VAMP calculator to determine if you are in Early Warning (0.9%) or Excessive (0.65%) territory before receiving an enforcement notice.
4. Review Mastercard subscription billing compliance before July 2026. Ensure pre-charge notifications are in place for all recurring billing. Audit your SMMP exposure if you have elevated scam-coded disputes.
5. Segment your dispute response strategy by fraud type. First-party fraud, third-party fraud, and agentic commerce disputes require different evidence packages and representment approaches. A single generic response template underperforms across all categories.
6. Audit BNPL dispute handling with your BNPL provider. Clarify who absorbs BNPL chargebacks, what the evidence requirements are, and what timelines apply for each BNPL provider you work with.

Frequently Asked Questions