Broadcom (AVGO) launches Arcot Scam Defense as AI-powered payment fraud challenges bank security

Broadcom Inc. (NASDAQ: AVGO), the US semiconductor and infrastructure software company, has launched Arcot Scam Defense, a payment security capability designed to identify social engineering fraud during card-not-present transactions even when a cardholder successfully completes an authentication challenge. Announced on October 21, 2026, the feature analyses transaction details, device intelligence and authentication behaviour to detect suspicious patterns that conventional verification systems may not identify. Broadcom said a pilot involving one major card issuer indicated that examining the highest-risk 1% of successful authentication challenges could have prevented an estimated $4 million in fraud during a single month. The launch expands the company’s payment security capabilities, although its wider commercial significance will depend on customer adoption, sustained fraud detection performance and the ability to limit unnecessary interference with legitimate transactions.

Why are banks struggling to detect payment scams even after customers complete security authentication?

Financial institutions commonly use additional authentication steps to establish that the person approving an online payment is authorised to use the account. These checks can include one-time passwords, biometric verification or approval through a banking application. The process can substantially reduce unauthorised transactions, but it does not necessarily determine whether the account holder has been manipulated into approving a fraudulent payment.

Social engineering attacks exploit that limitation by persuading customers to complete legitimate security steps for transactions that benefit criminals. A fraudster may impersonate a trusted organisation or create a false sense of urgency, encouraging the victim to approve an authentication request. The payment system can then receive apparently valid confirmation even though the customer’s decision was influenced by deception.

Broadcom’s latest product is designed to examine the circumstances surrounding the authentication rather than relying solely on whether the challenge was completed successfully. That approach addresses a category of fraud in which the legitimate account holder becomes an unwitting participant. The company is attempting to identify behavioural inconsistencies that may indicate manipulation despite otherwise valid credentials.

The commercial need for improved detection is linked to increasingly sophisticated attacks using impersonation and artificial intelligence-assisted communication. Fraudsters can imitate familiar brands, generate convincing messages and guide victims through legitimate payment approval processes. These techniques create risks that cannot always be addressed simply by introducing another password or authentication prompt.

Arcot Scam Defense targets card-not-present transactions processed through the 3-D Secure authentication environment. It is not a universal system covering every form of bank transfer, cryptocurrency transaction or account takeover. Its performance must therefore be evaluated within the payment workflows and fraud categories it is designed to protect.

How does Broadcom Arcot Scam Defense identify suspicious payments that appear successfully authenticated?

Arcot Scam Defense evaluates multiple contextual signals during the transaction and authentication process. These include payment information, device characteristics, historical behaviour and the manner in which a customer completes the security challenge. The system combines the signals to inform a real-time fraud decision instead of relying on any individual indicator.

One example involves differences between the device used to initiate a transaction and the device used to approve it. Large geographical separation or unusual device behaviour may indicate that the transaction is being conducted by someone other than the account holder. However, such differences can also occur during legitimate activity, making broader contextual evaluation necessary.

The system can also examine delays in entering one-time passwords or unexpected changes in authentication methods. These behaviours may occur when a fraudster is communicating instructions to a customer during the payment process. Their value as indicators depends on how reliably the system distinguishes suspicious activity from ordinary customer behaviour.

Broadcom has also described the potential use of third-party signals indicating whether the authenticating device is engaged in an active phone call. Such information may be relevant when a victim is receiving instructions from a fraudster while approving a payment. Access to these signals depends on the availability and integration of the relevant third-party capabilities.

The resulting risk assessment can allow financial institutions to challenge or block transactions exhibiting suspicious combinations of behaviour. However, Broadcom has not disclosed the complete decision model or independent detection results across a representative population of banks. The product’s effectiveness will depend on its performance in actual operating environments rather than the sophistication of its proposed analytical approach alone.

How should banks interpret Broadcom’s claim that Arcot Scam Defense could have prevented $4 million in fraud?

Broadcom reported that pilot testing using a full month of live transaction data produced encouraging results for one major card issuer. Its analysis indicated that reviewing the highest-risk 1% of successful authentication challenges could have protected against an estimated $4 million in fraud during that period. The company said the associated reduction in authentication success rate would have been approximately 0.38 percentage points.

These figures describe a pilot estimate, not independently established savings realised across the wider financial services industry. The amount relates to a particular issuer and transaction population, and the announcement does not disclose enough information to determine whether other banks would experience similar results. Differences in customer behaviour, fraud patterns and existing controls can materially influence performance.

The trade-off between fraud reduction and authentication success is particularly important. Blocking more transactions can reduce certain losses, but it may also interrupt legitimate purchases and create friction for customers. Banks must determine whether the value of prevented fraud justifies the operational costs, lost transactions and customer dissatisfaction associated with incorrect decisions.

Broadcom also reported nearly $500 million in fraud over the preceding twelve months involving successfully authenticated challenges within the activity underlying its disclosure. The figure illustrates the scale of the problem identified in that context but should not be represented as a comprehensive measure of global payment fraud. The company has not supplied a methodology establishing its applicability to every card issuer or payment network.

The commercial proposition is strongest where the system identifies high-risk transactions that existing authentication tools would otherwise approve. Financial institutions could potentially incorporate the additional capability without requiring every customer to complete more verification steps. However, operational outcomes will depend on how banks configure risk thresholds and integrate the product into existing decision systems.

Could improved behavioural fraud detection create new revenue opportunities for Broadcom’s infrastructure software business?

Broadcom operates a substantial infrastructure software business alongside its semiconductor operations. In the third quarter of fiscal 2026, infrastructure software generated approximately $8.75 billion in revenue, representing 30% of consolidated sales. Semiconductor solutions contributed approximately $20.84 billion, accounting for the remaining 70%.

The infrastructure software division provides technologies serving enterprise computing, security and other critical applications. Arcot Scam Defense extends the capabilities of an existing payment authentication offering rather than introducing an entirely new corporate business segment. The launch therefore represents a potential opportunity to deepen customer relationships within financial services.

Banks and payment issuers may be willing to invest in technologies that reduce fraud losses without materially reducing legitimate transaction approvals. Successful fraud prevention can produce measurable financial value, particularly when losses, investigation expenses and customer support costs are substantial. Broadcom could potentially use these outcomes to strengthen adoption of its payment security services.

However, the announcement does not disclose separate pricing, sales commitments or expected revenue attributable to Arcot Scam Defense. It would therefore be inappropriate to forecast a specific contribution to Broadcom’s earnings based on the pilot findings alone. The financial effect will depend on product adoption, contractual arrangements and whether customers purchase incremental capabilities.

The company has indicated that it intends to expand the product’s analytical capabilities through additional signals and improved historical profiling of authentication behaviour. Such development could increase detection accuracy if the new information provides useful distinctions between legitimate and fraudulent activity. It may also introduce further requirements relating to data integration and customer privacy controls.

How does Broadcom’s payment security initiative fit its wider financial performance and stock valuation?

Broadcom reported third-quarter fiscal 2026 revenue of $29.59 billion, representing growth of 86% from the corresponding period last year. The company’s AI semiconductor revenue reached $16.7 billion, increasing 221% year on year, while infrastructure software revenue grew 29%. These figures demonstrate that the group’s overall financial performance is currently influenced heavily by AI infrastructure and semiconductor demand.

Operating cash flow reached approximately $14.2 billion during the quarter, while free cash flow totalled $13.7 billion. The company also reported cash and cash equivalents of approximately $24 billion at the quarter’s end. This financial scale provides capacity to support software investment, but the October product announcement does not establish a material change in consolidated earnings expectations.

Broadcom shares closed at $321.68 on October 23, 2026, declining approximately 0.48% during the session. The stock’s reported 52-week range extended from $289.96 to $495, reflecting substantial volatility in a company valued at approximately $1.54 trillion. Market sentiment has been influenced principally by expectations for artificial intelligence semiconductors and wider technology spending.

Arcot Scam Defense represents a different commercial opportunity from the company’s fast-growing AI chip business. Its significance lies in strengthening an established software offering and addressing a specific problem faced by financial institutions. The product may improve competitiveness and customer retention without becoming a major standalone contributor to group revenue.

The latest launch expands the types of payment risk Broadcom can help institutions assess, particularly where customers complete legitimate authentication after being deceived. The next meaningful evidence will come from broader deployment results showing how many fraudulent transactions are prevented, how often legitimate purchases are interrupted and whether issuers adopt the capability commercially. Strong performance across those measures would establish a more persuasive business case than an individual pilot estimate.

Key takeaways from Broadcom Arcot Scam Defense and the growth of social engineering payment fraud

• Broadcom launched Arcot Scam Defense on October 21, 2026.
• The product targets social engineering during card-not-present payment authentication.
• It evaluates transaction, device and behavioural information.
• The system is designed to identify suspicious activity even after successful authentication.
• A pilot indicated an estimated $4 million in potentially preventable fraud during one month.
• The pilot involved reviewing the highest-risk 1% of successful authentication challenges.
• The reported authentication success-rate impact was approximately 0.38 percentage points.
• Broadcom’s infrastructure software division generated $8.75 billion in its latest completed quarter.
• No separate revenue forecast for Arcot Scam Defense has been disclosed.
• Wider bank adoption and verified fraud reduction will determine commercial significance.