The intellectual property strategies of payment giants Mastercard and Visa are diverging significantly, according to recent patent publication analysis. Mastercard is actively expanding its innovation agenda, with a notable increase in patent filings related to agent-initiated payments, open banking, and quantum security. In stark contrast, Visa’s patent portfolio remains heavily concentrated in its traditional domains of authorization and fraud prevention. This divergence is occurring even as the overall volume of patent publications within the payments sector has seen a decline, suggesting Mastercard is not merely maintaining filing volume but is strategically broadening its research and development focus into emerging technological frontiers.
This shift in focus has been observed over the ten months leading up to July 2026, during which patent publications across a select group of major payment industry players fell by 14.1% compared to the same period in the previous year. The analyzed cohort, chosen to represent diverse business models within the U.S. payments ecosystem, includes card networks (Mastercard and Visa), a card issuer (Capital One), a universal bank (JPMorgan Chase), and two prominent fintechs (PayPal and Block). It is important to note that this selection is not an exhaustive ranking of the largest patent filers in the sector, as other significant entities such as American Express, Fiserv, FIS, and Global Payments were not included in this specific analysis. The methodology of counting publications by their publication date ensures a like-for-like comparison across periods.
The contraction in patent publications is largely attributed to two major banking institutions. Capital One and JPMorgan Chase together account for 73% of the overall decline. Capital One, in particular, saw its own patent output decrease by 41% in the current year, a reduction that is substantial enough to represent more than the combined decrease from the two major card networks. This contraction appears to be concentrated within large, in-house bank portfolios rather than emanating from the card networks themselves. It is also crucial to acknowledge that publication counts are subject to the throughput of patent offices and do not necessarily reflect a company’s decision to cease innovation. Therefore, the data is most valuable when used for comparative analysis between companies rather than as an absolute measure of innovation activity.
Figure 1: Only Mastercard and Block Increased Patent Publications; The Other Four Fell by 15% or More
Source: GlobalData Patent Analytics
AI Language is Rising While AI Classifications Shrink
A more granular examination of patent publication titles reveals a fascinating trend: an increasing use of explicit generative AI and AI-agent terminology. Publications referencing terms such as "large language models," "foundation models," "prompts," "multi-agent frameworks," and "agentic systems" have surged from just 31 in the prior period to 104. This represents a significant jump in their share of total output, moving from a modest 0.6% to a more substantial 2.4%.
However, this upward trend in specific terminology is somewhat contradicted when looking at broader AI classification data. Publications carrying any general artificial intelligence or machine learning classification have actually decreased by 24.4%. Even the subgroup specifically related to generative models remained flat. This indicates that while companies are increasingly willing to use specific generative AI and agentic language in their public disclosures, the overall volume of patents being classified under broad AI categories is shrinking. This suggests a strategic re-framing of innovation disclosures rather than an exponential increase in the underlying AI development itself. Companies appear to be more precise in how they publicly describe their cutting-edge work, signaling intent rather than a simple expansion of scale.
This shift in descriptive language is not evenly distributed. Generative and agent disclosures now constitute a significant portion of PayPal’s publications, making up 7.4% of its total output, followed by Mastercard at 5.0%, and JPMorgan Chase at 3.4%. These figures represent a notable increase from the previous year, where they stood at 1.6%, 0.7%, and 1.1% respectively. Capital One’s generative and agent disclosures stand at 1.0%, while Visa shows a more modest increase, moving from 0.5% to 0.7% of its publications. Block, the only other company to experience an increase in its overall publication count, reported no disclosures falling into these specific generative AI and agent categories.
Figure 2: Mastercard’s Generative AI and Agent Disclosures Climbed to 5.0% of its Publications While Visa’s Barely Moved
Source: GlobalData Patent Analytics
Mastercard Added Clusters Without Losing Volume
Mastercard stands out as the sole entity within the analyzed group to achieve growth in both its overall patent output and its diversification into new technological areas. The company’s patent publications saw a modest increase of 2.2%. When patents acquired through strategic acquisitions, such as NuData Security, Vocalink, and CipherTrace, are excluded, this growth becomes even more pronounced, rising to 6.3%.
The company has notably established entirely new clusters of innovation. Previously, there were no patent publications specifically referencing "open banking." In the latest period, 25 such publications emerged, accounting for 4.5% of Mastercard’s total output. Intriguingly, many of these open banking patents also integrate concepts related to large language models, hinting at a convergence of these technologies.
Furthermore, Mastercard has made significant strides in quantum-resistant technologies. Publications related to quantum security have increased from a single instance to nine. These include critical advancements such as quantum-resistant offline authentication for payment devices, addressing a future threat to current cryptographic standards.
Perhaps most significantly, Mastercard has filed a patent application titled "Agentic microtransactions through verifiable payment delegation." This disclosure is unique across all 9,403 publications reviewed in both periods, being the only one to explicitly use the term "agentic" in its title. This highlights Mastercard’s proactive engagement with the concept of autonomous agents facilitating payments.
Mastercard is also strategically expanding its global intellectual property protection. The proportion of its patent filings through the World Intellectual Property Organization (WIPO) and the European Patent Office (EPO) has increased from 27.7% to 33.3% of its total, while its share of U.S. filings has decreased from 59.4% to 53.5%. This indicates a deliberate effort to secure its innovations on a broader international stage.
These emerging patent clusters are primarily concentrated in areas such as open banking, data services, and agent-initiated payment authorization. These are precisely the domains where future commercial applications and revenue streams are likely to materialize first, suggesting a clear strategic intent behind Mastercard’s innovation investment.
Visa’s Mix Did Not Move
In contrast to Mastercard’s expansionary approach, Visa has experienced a decline in its patent publication volume. The company’s publications fell by 18.4%. Even after excluding patents acquired through the Featurespace acquisition, the contraction remains significant at 16.2%. The decline is particularly sharp in U.S. publications, which are down by 27.8%.
Crucially, the composition of Visa’s patent portfolio has remained remarkably consistent. Approximately two in five of Visa’s AI-classified publications are still concentrated in the areas of transaction authorization, security, or fraud prevention. This ratio has barely shifted, standing at 40.4% this year compared to 40.2% in the previous year. Its generative model classifications have seen a slight increase, moving from six publications to nine, with one disclosure prioritizing 2025 focusing on foundation model fine-tuning. Overall, the patent landscape disclosed by Visa this year appears to be a smaller version of its portfolio from the previous year, with little indication of a strategic pivot.
It is important to clarify that Mastercard has not abandoned its traditional strengths in risk management. In fact, it dedicates a higher proportion of its AI-related patents to authorization and security than Visa does, at 53.4%. The key differentiator is that Mastercard has actively added new innovation clusters alongside its existing strengths, whereas Visa has not demonstrated a similar expansionary strategy in its patent disclosures.
Blockchain and Cryptocurrency Patent Activity Declines
An interesting observation that runs counter to some prevailing narratives is the decline in blockchain and cryptocurrency-tagged publications. Across the entire group analyzed, these publications fell by 27.3%. The two card networks, Mastercard and Visa, account for a significant 73% of this decline. Mastercard’s own blockchain-tagged output has decreased by 34.9%, and Visa’s by 36.5%. Across more than 9,000 publications reviewed, not a single title contained the word "stablecoin," even in the period following the establishment of a federal stablecoin framework by the GENIUS Act in July 2025. This suggests a potential cooling of interest or a strategic shift away from publicly patenting in this specific domain by these major players, at least in the immediate aftermath of regulatory clarity.
What the Data Cannot Settle
It is imperative to understand the limitations of this analysis. The data presented pertains to patent publications, not patent filings. A significant majority (72%) of the "frontier" publications, those in emerging technology areas, carry priority dates from 2024. Filings made after the significant announcements regarding agentic commerce in the spring of 2025 will not appear in publication volumes until late 2026 and throughout 2027. Therefore, the current analysis based on title keywords represents a minimum baseline of innovation activity rather than a complete census.
Three key developments will be crucial in testing these observations. The publication windows for late 2026 and 2027 will be the first to include filings with 2025 priority dates. These will provide the initial concrete evidence regarding whether Visa has indeed patented its work in agentic commerce or if this innovation remains unpatented. Mastercard’s patent applications have been on the rise while its granted patents have seen a decline, indicating a growing pipeline of pending applications. The next 12 to 24 months will reveal whether these applications translate into granted patents and commercial products. Furthermore, any emergence of "stablecoin" or "tokenized settlement" language in future publications would provide a clear indication of the industry’s real-time filing response to the GENIUS Act and its regulatory framework.
Patent Activity is One Thing. Whether it is Investable is Another.
The analysis of patent activity, while informative about innovation direction, is only one facet of evaluating a company’s future prospects. Whether this patent activity translates into investable opportunities is a separate, albeit related, question. GlobalData’s proprietary research, such as their report "Extracting Innovation Alpha Using Patents," delves into this very subject. By analyzing over 40 unique patent indicators across the S&P 500 from 2019 to 2025, their research has demonstrated that patent-identified leaders can generate significant alpha. The findings suggest annualized returns of between 5% and 7% above the S&P 500, and between 6% and 9% above benchmarks like the ARK Innovation ETF and the NYLI US Large Cap R&D Leaders Index. Their generative AI leadership matrix, for instance, showed that patent-identified leaders in this field outperformed laggards by a factor of 1.7. The Brinson attribution model further helps to separate the impact of stock selection from sector allocation in these performance metrics. Understanding these nuances is critical for investors seeking to capitalize on technological innovation.
