Fraud is getting harder to distinguish from legitimate digital activity. AI-generated scams blend into social feeds and messaging apps, while high-traffic events give fraudsters more opportunities to hide among legitimate customers. Even when overall fraud rates decline, coordinated attacks can target individual businesses without drawing attention.
These were some of the challenges explored in Sift’s recent webinar, AI Scams, Peak Events, and the Cost of a Slow Response. Sift Trust and Safety Architects Maria Benjamin and Jacob Sanchez discussed findings from the Q3 2026 Digital Trust Index, combining a survey of more than 1,000 consumers with fraud trends from the Sift Global Data Network.
Here are five insights for fraud teams looking to manage emerging threats, protect customer trust, and prepare for peak events.
Consumers are struggling to recognize AI-generated scams
More than half (53%) of consumers have encountered an AI-generated scam on social media or messaging apps. Yet only 36% feel very or extremely confident they could identify one.
Jacob pointed to the viral AI-generated videos of Will Smith eating spaghetti as an example of how quickly the technology has advanced. A few years ago, the videos were obviously fake. Today’s AI-generated content is far more convincing, making fraudulent advertisements, images, and videos increasingly difficult to recognize.
“The tells are getting covered by advancements in AI generation,” Jacob explained.
Fraud tactics are also becoming more accessible. The Digital Trust Index found that 37% of consumers have encountered social media content teaching or selling fraud methods, while 24% have tried a social media “hack” to obtain a refund, discount, or free item.
Jacob shared an example involving food delivery: using AI to alter a photo of a meal so it appears raw, then submitting it to request a refund. Because these tactics are presented as harmless ways to save money, consumers may not recognize them as fraud.
For businesses, this creates new challenges in detecting scams, refund abuse, and manipulated evidence while protecting legitimate customers.
Security friction can strengthen customer trust
Consumers are increasingly taking steps to protect themselves online. The Digital Trust Index found that 62% regularly monitor bank or card statements, 60% avoid links from social media or messaging apps, and 56% use two-factor or multi-factor authentication.
These behaviors reflect growing awareness of online threats, but they also influence what consumers expect from businesses. Jacob noted that he’s less likely to sign up for an app that lacks basic security protections. Maria described how features such as two-factor authentication can actually increase customer confidence.
“Friction becomes a trust mechanism here, where a lot of people are like, if I can’t actually enable two-factor authentication, I’m out,” Maria explained.
For fraud teams, the challenge is determining where friction adds value and where it unnecessarily disrupts legitimate activity. Customers may expect a fast checkout experience but appreciate additional verification before a large withdrawal.
Applying risk-based friction at the right moments can help businesses strengthen security while keeping routine interactions straightforward.
The cost of fraud goes beyond the transaction value
Fraud rates don’t always reflect a business’s full financial exposure. Across the Sift Global Data Network, average order value increased 52% year over year to $128. That means businesses can face higher fraud losses even when the percentage of fraudulent transactions stays the same.
Businesses also need to consider the cost of incorrectly blocking legitimate customers. In iGaming, a false positive costs an average of $335, compared with an average order value of $216. In e-commerce, those figures are $292 and $108, respectively.
This is particularly significant in iGaming, where operators invest heavily in customer acquisition through advertising, sign-up bonuses, and promotions. Incorrectly blocking a legitimate player can mean losing that investment along with potential future revenue.
Maria and Jacob discussed how layered fraud controls can help manage this tradeoff. An operator might keep registration relatively simple, then introduce stronger checks when a player attempts to withdraw funds. Customers may even expect additional security at that stage, particularly for larger amounts.
For fraud teams, the priority is balancing fraud prevention with the cost of unnecessary declines, applying stronger controls where the risk justifies them.
A strong fraud response can help rebuild customer trust
A fraud incident can affect a customer relationship long after the initial financial loss. According to the Digital Trust Index, 47% of consumers have stopped using or reduced their use of a company following a fraud experience.
But how a business responds can significantly influence whether customers return. The research found that 73% of consumers respond positively when a company clearly explains what went wrong. Proactive communication improves perception for 71%, while resolving an incident within a week improves it for 63%.
Maria emphasized the importance of keeping customers informed about what happened, how the business is addressing it, and when they can expect a resolution. Jacob added that companies take different approaches to account takeover recovery. Some place responsibility on the customer, while others offer additional support or incentives to restore the relationship.
“A recovery experience that’s great, customer-minded, can create a promoter even when there’s something bad happening to them,” Jacob said.
For fraud teams, incident response is also an opportunity to protect customer retention. Clear communication and timely resolution can help restore confidence after something goes wrong.
Peak events create more opportunities for fraud to hide
Peak shopping periods bring increased transaction volume and revenue opportunities, but they also give fraudsters more opportunities to blend attacks into legitimate traffic.
In Sift’s consumer survey, 64% reported spending the most during the holiday and New Year season. However, Maria and Jacob noted that peak events vary considerably by industry. Food delivery businesses may see spikes around back-to-school, while restaurants experience increased demand around Valentine’s Day and Mother’s Day. Major sporting events can also drive activity across grocery, retail, and food delivery.
The full Digital Trust Index illustrates why businesses need to look beyond seasonal trends and network-wide fraud rates. Although payment fraud attack rates declined 14% year over year, anomaly monitoring across nine merchants identified 129 critical anomalies over six weeks. Of those, 82% matched BIN-attack patterns. One software platform experienced suspicious activity for more than 41 consecutive days without an obvious seasonal trigger.
These findings reinforce the importance of continuous monitoring, even when overall fraud trends appear favorable.
How to prepare for the next peak event
A strong fraud strategy starts before transaction volumes increase and continues after the event ends.
- Before the event: Identify high-risk customer journeys, establish normal transaction patterns, align teams on escalation procedures, and test fraud controls before traffic increases.
- During the event: Monitor approval rates, block rates, review queues, and transaction velocity. Look for unusual patterns and adjust controls as risks change without slowing trusted customers.
- After the event: Review fraud losses, false positives, customer friction, and response times. Monitor delayed disputes and chargebacks, then apply those findings to future planning.
Prepare for fraud beyond the next peak event
Sift’s Q3 2026 Digital Trust Index shows how fraud is evolving across the customer journey. AI-generated scams are becoming harder to recognize, fraud and false positives carry growing financial consequences, and customer trust depends on how businesses respond when something goes wrong.
For fraud teams, effective preparation means understanding normal transaction behavior, identifying unusual activity early, applying friction where it matters, and responding quickly to emerging threats.
Every business has its own peak events, traffic patterns, and risk tolerance. Its fraud strategy should reflect those differences.
Download Sift’s Peak Event Fraud Readiness Checklist for practical steps to prepare, monitor, and respond.
For more insights, read the Q3 2026 Digital Trust Index or watch the full webinar recording.
