J.P. Morgan Growth Equity Partners has taken a strategic stake in Island, the Israeli-US company behind the “enterprise browser,” becoming the latest blue-chip backer in the start-up’s $250 million Series E financing at a valuation of roughly $5 billion, Island said on Monday. The investment, made through J.P. Morgan’s $1 billion growth fund, follows the March round led by Coatue Management and brings Island’s total capital raised to about $730 million. Other investors include Sequoia Capital, Insight Partners, and Cyberstarts. “Cybersecurity is at the top of the priority list for the world’s largest organizations, and Island is exactly the type of company we aim to support,” Paris Heymann, co-managing partner at J.P. Morgan Growth Equity Partners, said in the release. The arrival of the Wall Street bank “is testimony to the value we bring,” added Island co-founder and CTO Dan Amiga, noting that “many of the world’s largest banks have already chosen Island.” Launched from stealth only in February 2022, Island says its secure browser now runs at more than 450 enterprises, including eight of the 10 largest US banks. The software lets corporate IT teams bake security controls, data-loss prevention and productivity tools directly into the browser layer, making remote work and bring-your-own-device policies easier to police. The company employs roughly 500 people—about 200 of whom are based at its R&D hub in Tel Aviv—and is co-headquartered in Dallas. The company was founded in 2020 by Amiga, a former Unit 8200 officer and serial entrepreneur, and CEO Mike Fey, the onetime president of Symantec and CTO of McAfee. For Island, the J.P. Morgan check delivers both fresh capital and a marquee customer reference in the financial services sector—a vertical that has accounted for the company’s earliest adopters and remains its fastest-growing segment. Heymann’s fund has been making late-stage bets on enterprise software and cybersecurity since its launch last year; its move on Island “underscores how critical browser-level security has become to regulated industries,” the company said. With the new money, Island plans to accelerate hiring in engineering and go-to-market roles, expand its Dallas and Tel Aviv sites, and double down on product integrations aimed at large financial institutions.
Capital One’s Head of Enterprise AI opines bank CIOs with an obsession for calculating ROI in AI projects “are making irreversibly bad decisions” in an environment of dramatically changing costs of inference
Prem Natarajan, EVP and head of enterprise AI at Capital One, says the economics of gen AI costs are changing so dramatically that attempts to use traditional financial tools to calculate and project AI ROI is the wrong way to go about it. “In the last 22 months, the cost of inference has come down by more than a factor of 1,000 on a performance equivalent basis,” Natarajan tells CIO.com. “Something that cost you $10 to do inference on [two years ago] is now costing you one cent. In that environment of dramatically changing costs, any focus on near-term robust prediction of ROI as a justification for investing in gen AI” is likely to fail. Natarajan, who described the economic changes as being “on the throes of a generational inflection,” believes that CIOs taking that approach to projecting ROI “are making irreversibly bad decisions that will make them fall behind” given their “obsession of calculating ROI in the face of transformative technologies.” Natarajan joined Capital One in March 2023 from Amazon, where he spent almost five years as vice president for Alexa AI. He estimates that at Capital One he oversees “several hundred petabytes of data that will approach exabyte scale” in “a couple of years.” That data trove is a key asset for Capital One in making the most of AI, according to Natarajan, who sees data governance and accessibility as additional keys to AI success. Jason Andersen, a vice president and principal analyst tracking AI for Moor Insights & Strategy, says Natarajan’s take on the ROI issue for CIOs is valid. “Enterprises [such as Capital One] are starting to get really smart about how they are deploying AI and building AI applications,” Andersen says. “The reality is that we haven’t seen a trend in technology ever move this fast — ever.” That speed has caught many IT executives off guard as techniques that have always worked for them stop working, Andersen adds. “With this absolute velocity, you are seeing the old norms of trying to figure out how much to invest, those are no longer useful tools,” he says. “If you use traditional methods, you just don’t get it.” With almost any form of AI, he says, “your data advantage is your AI and ML advantage.” “The amount of proprietary data we had was an important asset to be brought to life in building generative AI applications and capabilities that would be differentiating for us,” Natarajan says, stressing that their evaluations showed that they “could not use closed-source models, because you cannot meaningfully customize those models.” The Capital One AI team eventually opted to use Meta’s open-source Llama LLM and set about building AI solutions atop its public cloud foundation, established prior to Natarajan’s tenure at the financial services company. “Capital One was the first bank — and to date, the only bank — that is all in on the public cloud. They shut down all their data centers over a period of two years and moved everything to AWS,” Natarajan explains. “We became cloud-native developers.” At Capital One, which employs roughly 14,000 IT specialists, talent is critical but so too is data — perhaps more so, Natarajan says.
AWS model distillation feature transfers intelligence from a larger model to a smaller, more specialized model by generating 10X more synthetic data based on customer prompts
AWS is preparing for the upcoming AWS re:Invent later this year, with a series of product updates based around intelligent automation and agentic AI. “We are seeing employability for some of these cloud models. We also have been busy launching some of our first party models with Nova,” said Atul Deo, director of product, AWS Bedrock, at AWS. Since announcing a new generation of foundation models at the last re:Invent, AWS has made intelligent prompt routing generally available. This tool enables users to combine the advantages of cheaper and larger, more capable models. Another product that offers the best of both worlds is Bedrock’s model distillation feature, which transfers intelligence from a larger model to a smaller, more specialized model. “We’ll generate additional data for the distillation process based on the prompts that a customer provides,” said Deo. “It can give a few indicator 30, 40 prompts of what it wants kind of generally for the distillation purpose. Then behind the scenes we can generate 10 times more data, which is basically synthetic data, then that synthetic data response of that larger model then gets used to essentially kind of make the smaller model more targeted and focused.” Two of the hottest areas for generative AI have been code generation and sales and marketing. Part of making AI a good assistant for customer service, code or even real estate is providing agents standardized access to relevant context through Model Context Protocol, according to Deo.
New energy-based transformer (EBT) model architecture enables building cost-effective AI applications that can generalize to novel situations without the need for specialized fine-tuned models
Researchers at the University of Illinois Urbana-Champaign and the University of Virginia have developed a new model architecture that could lead to more robust AI systems with more powerful reasoning capabilities. Called an energy-based transformer (EBT), the architecture shows a natural ability to use inference-time scaling to solve complex problems. For the enterprise, this could translate into cost-effective AI applications that can generalize to novel situations without the need for specialized fine-tuned models. EBTs are trained to first verify the compatibility between a context and a prediction, then refine predictions until they find the lowest-energy (most compatible) output. This process effectively simulates a thinking process for every prediction. The researchers developed two EBT variants: A decoder-only model inspired by the GPT architecture, and a bidirectional model similar to BERT. The architecture of EBTs make them flexible and compatible with various inference-time scaling techniques. Crucially, the study found that EBTs generalize better than the other architectures. Even with the same or worse pretraining performance, EBTs outperformed existing models on downstream tasks. The benefits of EBTs are important for two reasons. First, they suggest that at the massive scale of today’s foundation models, EBTs could significantly outperform the classic transformer architecture used in LLMs. Second, EBTs show much better data efficiency. This is a critical advantage in an era where high-quality training data is becoming a major bottleneck for scaling AI.
US regulators provide blueprint for lenders’ crypto custody – requiring a risk-governance framework that appropriately adapts to relevant risks
US regulators gave fresh guidelines for how banks can offer crypto custody services and not run afoul of rules. Banks that contemplate providing safekeeping for crypto-assets should consider the evolving nature of the crypto market, including the technology underlying the cryptoassets, regulators said. The Federal Reserve, Federal Deposit Insurance Corp. and Office of the Comptroller of the Currency said firms must also implement a risk-governance framework that appropriately adapts to relevant risks. The statement outlined key risk areas and warnings for banks to consider: Potential risks prior to offering crypto safekeeping; Being held liable for customers’ losses in cases of possible compromise or loss of cryptographic keys or other sensitive information; Crypto safekeeping relationships are subject to applicable Bank Secrecy Act/Anti-Money Laundering laws; Risks from contracting with a third-party; Appropriate audit coverage — especially assessing management and staff expertise.
JPMorgan has established a new unit within its commercial and investment bank focused on creating bespoke financing structures that span public and private markets
JPMorgan Chase & Co. has established a new unit within its commercial and investment bank focused on creating bespoke financing structures that span public and private markets. Named Strategic Financing Solutions, the unit will integrate efforts across banking, markets, and sales, according to an internal memo circulated Monday. Initially, the unit will concentrate on structured private solutions, infrastructure finance, strategic asset-backed securities finance, merchant banking, and direct lending. “Financing needs are becoming increasingly complex while investors are seeking exposure to new markets,” wrote Doug Petno and Troy Rohrbaugh, co-heads of the commercial and investment bank. The memo stated that the new team “will focus on delivering alternative solutions to our corporate and sponsor clients.” Warfield Price, head of general industries leveraged finance, and Masi Yamada, global head of corporate structuring, will co-lead the group while retaining their current roles. They will report to Kevin Foley, global head of capital markets, and Brad Tully, global head of private side sales and corporate derivatives. The initiative targets the convergence of public and private markets, noting that many U.S. firms are choosing to remain private longer and are demanding more complex financial solutions. A notable example is JPMorgan’s advisory role in 3G Capital’s $9.4 billion acquisition of Skechers USA Inc., a deal involving cash, debt, two term loans, two notes, and a revolving credit facility. The new unit will also encompass JPMorgan’s direct lending operations, for which the bank allocated an additional $50 billion this year after deploying over $10 billion across 100 deals since 2021. In a February interview, Foley emphasized the bank’s “product-agnostic” approach to serving clients. The move mirrors Goldman Sachs Group Inc.’s recent formation of a capital solutions group, which also reflects the rising significance of private markets.
JPMorgan 2Q’25 reports active mobile customers up 8% YoY; Consumer & Community Banking revenue up 6% YoY, predominantly driven by higher net interest income in card services
JPMorgan reported revenue was $44.9 billion and managed revenue was $45.7 billion. Expenses totaled $23.8 billion, resulting in a reported overhead ratio of 53% and a managed overhead ratio of 52%. Credit costs were $2.8 billion, including $2.4 billion of net charge-offs and a $439 million net reserve build. Average loans increased 5% year-over-year and 3% quarter-over-quarter, while average deposits rose 6% year-over-year and 3% quarter-over-quarter. In the Consumer & Community Banking (CCB) segment, return on equity (ROE) was 36%. Average deposits declined 1% year-over-year but rose 1% quarter-over-quarter; client investment assets grew 14% year-over-year. Average loans were up 1% year-over-year and flat quarter-over-quarter, with Card Services posting a net charge-off rate of 3.40%. Debit and credit card sales volume increased 7% year-over-year, and active mobile customers grew 8% year-over-year.
Consumer & Community Banking
Net income of $5.2B, up 23% YoY
Revenue of $18.8B, up 6% YoY, predominantly driven by higher net
interest income in Card Services on higher revolving balances, higher
noninterest revenue in Banking & Wealth Management, as well as higher
operating lease income in Auto
Expense of $9.9B, up 5% YoY, largely driven by higher technology
expense and higher auto lease depreciation
Credit costs of $2.1B
NCOs of $2.1B, up $22mm YoY, primarily driven by Card Services
Reserves were relatively flat, as changes in the weighted-average
macroeconomic outlook were offset by loan growth in Card Service
Commercial & Investment Bank
IB revenue of $2.7B, up 9% YoY, predominantly driven by higher debt underwriting and advisory fees, partially offset by lower equity underwriting fees
Payments revenue of $4.7B, up 4% YoY; excluding the net impact of equity investments, revenue up 3%, driven by higher deposit balances and fee growth, predominantly offset by deposit margin compression Lending revenue of $1.8B, down 6% YoY, largely driven by higher losses on hedges of the retained lending portfolio
Asset & Wealth Management
Revenue of $5.8B, up 10% YoY, driven by growth in management fees on strong net inflows and higher average market levels, as well as higher brokerage activity and higher deposit balances Expense of $3.7B, up 5% YoY, driven by higher compensation, includinghigher revenue-related compensation and continued growth in private banking advisor teams, as well as higher distribution fees AUM of $4.3T was up 18% YoY and client assets of $6.4T were up 19% YoY, each driven by continued net inflows and higher market levels For the quarter, AUM had long-term net inflows of $31B and liquidity net inflows of $5B Average loans of $241B, up 7% YoY and 3% QoQ Average deposits of $248B, up 9% YoY and 2% QoQ
Jamie Dimon, Chairman and CEO, commented on the financial results: “We reported another quarter of strong results, generating net income of $15.0 billion or net income of $14.2 billion excluding a significant item.” Dimon continued: “Each of the lines of business performed well. In the CIB, Markets revenue rose to $8.9 billion, and we supported clients as they navigated volatile market conditions at the beginning of the quarter. Meanwhile, IB activity started slow but gained momentum as market sentiment improved, and IB fees were up 7% for the quarter. In CCB, we added approximately 500,000 net new checking accounts, which drove sequential growth in checking account balances. In Card, we launched a refreshed Sapphire Reserve along with a new Sapphire Reserve for Business, with positive early reactions and strong new card acquisitions. Finally, in AWM, asset management fees rose 10%, and we saw continued client asset net inflows of $80 billion, with client assets crossing over $6.4 trillion.”
Citigroup 2Q’25- Wealth revenues were up 20% with growth across all three lines of business; in U.S. Personal Banking, good growth in Branded Cards while Retail Banking benefited from higher deposit spreads
Citigroup 2Q25 reported net income for the second quarter 2025 of $4.0 billion, or $1.96 per diluted share, on revenues of $21.7 billion. This compares to net income of $3.2 billion, or $1.52 per diluted share, on revenues of $20.0 billion for the second quarter 2024. Revenues increased 8% from the prior-year period on a reported basis, driven by growth in each of Citi’s five interconnected businesses, partially offset by a decline in All Other. Excluding divestiture-related impacts in both periods, revenues were up 9%. Net income was $4.0 billion, compared to $3.2 billion in the prior-year period, driven by the higher revenues, partially offset by higher cost of credit and higher expenses. Earnings per share of $1.96 increased from $1.52 per diluted share in the prior-year period, reflecting the higher net income and lower shares outstanding. Percentage comparisons throughout this press release are calculated for the second quarter 2025 versus the second quarter 2024, unless otherwise specified. Citi CEO Jane Fraser said, “We reported another very good quarter and continue to demonstrate that our strong results are sustainable through different environments. We’re improving the performance of each of our businesses to take share and drive higher returns. With revenue up 8%, Services continues to show why this high-return business is our crown jewel. Markets had its best second quarter performance since 2020 with a record second quarter for Equities. Banking revenues were up 18% and we continue to be at the center of some of the most significant transactions. Wealth revenues were up 20% with solid growth across all three lines of business. In U.S. Personal Banking, we saw good growth in Branded Cards while Retail Banking benefited from higher deposit spreads. We returned $3 billion in capital during the quarter, including $2 billion in share repurchases as part of our $20 billion repurchase plan. I’m particularly pleased that the momentum across our franchise includes the Transformation, as we streamline processes, drive automation and deploy AI. As I’ve said, next year’s 10–11% ROTCE target is a waypoint, not a destination. The actions we’ve taken have set up Citi to succeed long term, drive returns above that level and continue to create value for shareholders,” Ms. Fraser concluded. Citigroup Inc. today reported net income for the second quarter 2025 of $4.0 billion, or $1.96 per diluted share, on revenues of $21.7 billion. This compares to net income of $3.2 billion, or $1.52 per diluted share, on revenues of $20.0 billion for the second quarter 2024.
Wells Fargo 2Q’25 reports – Mobile active customers- 32.1 million in 2Q25 up from 31.8 million in 1Q25; Consumer Banking and Lending (CBL)- total revenue was up 2% year-over-year and up 4% from the first quarter of 2025
Wells Fatgo 2Q25 reported net interest income decreased 2%, driven by the impact of lower interest rates on floating rate assets and changes in deposit mix, partially offset by lower market funding and reduced deposit pricing. Noninterest income increased 4%, reflecting the gain from the merchant services joint venture acquisition, higher asset-based fees in Wealth and Investment Management due to improved market valuations, and an increase in investment banking fees, partially offset by lower net gains from trading in the Markets business
Consumer Banking and Lending (CBL)- Total revenue was up 2% year-over-year and up 4% from the first quarter of 2025. Consumer and Small Business Banking (CSBB) rose 3% year-over-year, driven by higher net interest income, and increased 5% from 1Q25 due to higher net interest income and seasonally higher debit card interchange income. Home Lending was down 5% from 1Q25, primarily reflecting lower mortgage servicing income resulting from portfolio run-off and sales. Credit Card revenue increased 9% year-over-year on higher loan balances. Auto declined 15% year-over-year due to lower loan balances and loan spread compression. Personal Lending was down 9% year-over-year, driven by lower loan balances. Mobile active customers- 32.1 million in 2Q25 up from 31.8 million in 1Q25
Commercial Banking (CB)- Total revenue was down 6% year-over-year but up modestly from the first quarter of 2025. Net interest income declined 13% year-over-year due to the impact of lower interest rates, partially offset by lower deposit pricing and higher deposit and loan balances. Noninterest income rose 13% year-over-year, driven by higher revenue from tax credit investments and increased treasury management fees. Noninterest expense increased 1% year-over-year as higher operating costs were partially offset by lower personnel expenses, reflecting the impact of efficiency initiatives; expenses were down 9% from 1Q25.
Chief Executive Officer Charlie Scharf commented, “Our second quarter results reflect the progress we are making to consistently produce stronger financial results with net income and diluted earnings per share up from both the first quarter and a year ago. Our efforts to increase fee-based income drove revenue growth and both net interest income and noninterest income grew from the first quarter. We are investing in our businesses but remain focused on expense management. While there continue to be risks as we look forward, activity levels have remained consistent and our strong credit performance continues to point to the strength of our commercial and consumer customers’ financial position.” Scharf continued, “The lifting of the asset cap in the second quarter marked a pivotal milestone in Wells Fargo’s ongoing transformation, along with the termination of thirteen consent orders since 2019, including seven this year alone. We are a far stronger company today because of the work we’ve done. This is a huge accomplishment, and I appreciate the focus and dedication that was required of everyone at Wells Fargo. We now have the opportunity to grow in ways we could not while the asset cap was in place and are able to move forward more aggressively to serve consumers, businesses, and communities to support U.S. economic growth.” He concluded, “As we have been investing to drive organic growth and improve the earnings capacity in each of our businesses, we have also been returning excess capital to shareholders. During the first half of this year, we repurchased over $6 billion of common stock and as previously announced, we expect to increase our third quarter common stock dividend by 12.5%, subject to approval by the Company’s Board of Directors at its regularly scheduled meeting later this month.”
JPMorgan Chase’s decision to charge data aggregators could accelerate the adoption of standards like FDX which aim to replace legacy methods like screen scraping with tokenized, API-based access.
While JPMorgan’s decision to charge for data access may not be unreasonable, it did catch many by surprise. The bank argues that aggregators are profiting from its infrastructure without contributing value in return. Citing rising infrastructure and security costs, as well as a desire for greater control over how consumer data is accessed and used, JPMorgan framed the move as a necessary step toward a more balanced data-sharing ecosystem. For data aggregators, the news is far from welcome. As one spokesperson noted, their cost of goods sold has essentially been zero. They charge fintechs for data access but haven’t had to pay banks to obtain the data itself. If banks like JPMorgan begin charging for that access, aggregators will likely pass the added costs to fintechs, which could ultimately trickle down to consumers. JPMorgan’s announcement comes at an interesting time for open banking in the US. Section 1033 of the Dodd Frank Act was supposed to be finalized this October, and many were looking forward to the clarity that centralized open banking rules would provide the industry. Earlier this year, however, the CFPB announced plans to rescind 1033. Regardless of whether or not formal rules are in place, however, the argument centralizes around an age-old question in fintech–who owns the customer data? While many banks claim that the consumer data belongs to them, some advocacy groups and aggregators claim that consumers should be able to do what they want with their data freely. Introducing new costs to access consumer financial data could have several ripple effects on the future of open banking in the US:
- It may create barriers for fintechs offering services that consumers can’t get from traditional banks. This could slow innovation and reduce incentives for new entrants to build products that meet unmet financial needs.
- Consumers may face higher costs as fintechs pass on the fees associated with data access. Services that were once free or low-cost could become more expensive, prompting some users to reconsider their primary financial institution if their bank can’t match the functionality they previously enjoyed via third-party apps.
- It could accelerate the adoption of more secure, standardized data-sharing protocols, such as those developed by the Financial Data Exchange (FDX), which aim to replace legacy methods like screen scraping with tokenized, API-based access.
- It might also incentivize more screen scraping, as aggregators seek ways to avoid new costs. While most aggregators treat screen scraping as a last resort, increased financial pressure may push some to lean more heavily on automated tools such as AI agents to extract data through less secure channels