- 2 days ago
Category
🗞
NewsTranscript
00:00We are back with a little bit more data and survey results and context, so I
00:07didn't get to introduce myself, no wonder, but I'm part of Bloomberg
00:11Intelligence. We are the in-house investment research team within Bloomberg.
00:16We are a team of 500 analysts, associates, and editors across all the regions in the
00:22world, including mainland China, including Brazil now. I just wanted to share
00:28something with you which is hot off the press, so this came out right this
00:31morning. We conducted a survey of 600 plus respondents across different industries
00:40to assess what was the real impact of AI in their industries, and there's a lot of
00:47hype around the AI bubble, around what AI can do. The biggest fear is, oh, will there
00:52be a lot of job losses, which industries are more vulnerable, blah blah blah. We
00:56wanted to really quantify that, and that's what my department does. We do
00:59data-based investment research. 150 participants out of these were from
01:05financial services, and that spans across insurance, asset and wealth
01:09management, commercial banking, investment banking, fintech, and payments. Most of this
01:15cohort, about 80% plus, was C-suites, with the remainder serving as senior directors or
01:22above, across these organizations. So this really is the real picture of what companies
01:27are thinking about AI adoption and where they are. To give you a sense of the size of
01:32the companies that we connected with, 50% plus were firms with 5,000 to 10,000 employees,
01:3830% of them were 10,000 to 50,000 employees, and 16% companies were over 50,000 employee base,
01:46and geographically, 40% were North America, 50% were Europe. So about 70% of financial institutions saw
01:561 to 10% revenue growth, while roughly 74% project a similar rise in profits. And if you think about it,
02:06this basically suggests very minimal operating leverage, if you have the same amount of translation
02:13of growth in revenue and profit. And what that means is that AI basically stands to benefit
02:20companies on client engagement, on personalization, on cross-selling revenue. Yet, because they're
02:28constantly investing in AI, in data technology, in infrastructure, in personnel, the impact on cost
02:35is going to be a bit more delayed. And that's what the data, the survey results are kind of pointing to.
02:40So we do expect cost ratios to start rationalizing a little bit more, as productivity gains materialize
02:47after that investment period is over, or at least the investments taper off. And that supports stronger
02:52incremental margin benefits beyond 2027-28 across the industry. More than 75% of our respondents
03:02expected productivity to rise at least 6%. Yet, over 70% also see higher operating costs because of AI
03:09deployment. And that's, as I mentioned, because of all the investments in data, cloud migration, AI-related
03:15tools, and model governance. The findings reinforce that AI's early phase in financial services is more
03:21likely to be about building capability rather than cost reduction.
03:27This is a very important chart for a lot of people, and this is something that a lot of people worry.
03:32Contrary to popular belief, two-thirds of financial services executives expect headcount
03:39to rise over the next three years due to AI. More than half of them are planning to add roles
03:45as AI compliance specialists, workflow, and agent designers, model auditors. So there are new roles that
03:52they anticipate creating because of all these investments in AI. And that basically signals a
03:58workflow realignment rather than a headcount reduction focus, and rather than basically displacing
04:04more people tied to this technology, right? Most foresee a 6% to 10% increase in net headcount. That
04:12is the other surprising takeaway from this survey. And that's likely concentrated in the areas around data
04:18science, engineering, research, and advisory roles. One thing to note, financial services in particular,
04:27the AI adoption curve is a lot more gradual than some of the other industries that you probably
04:33encounter, be it retail, consumer, tech. Only 23% of respondents actually report enterprise-wide
04:42scaling of AI tools compared to 35% across all industries. And the gap likely stems from overlapping
04:48regulatory data and risk constraints. There's tighter model governance, explainability requirements by
04:55regulators, fragmented legacy data as a big challenge across industry, and strict oversight of PII. Anyone
05:02who works in financial services here appreciates that challenge. All of these issues limit model
05:08training, and the strategies add cost and complexity to model development, whereas risk and compliance
05:14caution also slows out your ability to roll out models in real time and go at the pace at which other
05:20industries in tech or retail probably go. Operational efficiency, obviously, is one of the biggest
05:26focus, why people are investing in AI. And nearly half ranking, it has the top objective for their AI
05:33investments. Process automation is central, but firms are emphasizing productivity gains, not headcount cuts,
05:40with only 15% of these respondents citing job cuts as their primary goal. So people are wanting to do
05:47more with less rather than cutting down people in their firms at the moment. At the moment is the main
05:52point. Revenue generation and customer experience are following closely. That reflects a broader shift
05:57towards using AI to deepen more personalization across your product spectrum, accelerate product innovation,
06:05unlock new and digital revenue streams. If you talk to any banks, that's one of the key focus that they have.
06:09Here's the interesting part. I think the audience here will relate the most with this. 40% of financial
06:17services allocate 11 to 20% of their annual tech budgets to AI related initiatives. Payment companies
06:26across the financial spectrum lead that. An average of 17% of total spending is dedicated to AI capabilities.
06:35Investment banking and wealth management follow closely. Obviously leaders like Goldman Sachs,
06:39Morgan Stanley, BlackRock, they're all making very heavy AI investments.
06:45And here's the one of the most interesting pieces as well, right? In our view, AI could actually
06:51graduate from a cost focus to a margin and monetization lever. And we think that fintechs addressing SMBs
06:59are disproportionately positioned to benefit from AI. Why is that? Because all financial subsectors,
07:07they're actively pursuing AI adoption. But AI is one of the first technologies that can actually
07:13process unstructured workflows and information. And that has been one of the biggest challenges. And
07:19I'll talk more about it as we go forward. We expect winning fintechs to adapt AI as a lever to lower cost
07:26ratios, faster acquisition. However, companies with lower or thinner data modes, I think those are the
07:32companies that are going to struggle in the AI era. I'll delve a bit deeper on why the fintechs we think
07:39are better positioned. But just to kind of place the industry context, SMBs are the backbone of the
07:46global economy, right? They account for more than half of the global GDP, over 90% of all companies are SMBs,
07:52and more than two thirds of business employment is through SMBs. Every bank, what their salt would
07:59tell you what a lucrative segment this is, how much they want to target it. And yet, this is one of the
08:04biggest and most challenging segment that banks have struggled to service globally. Manual AR and AP
08:11processes dominate financial operations. There's just too much manual intervention, too much inefficiency
08:18that kind of is ridden. And that is what makes SMBs challenging. But from the client perspective,
08:24quality and cost of labor is what Mary Kay had pointed out during our panel. It's one of the most challenging
08:30parts of SMBs. And collectively, these pain points add to operating costs, they elongate working capital cycles
08:37for these companies, and they just keep it. It's like a chicken and egg problem that kind of keeps self reinforcing.
08:43AI is the first technology that is capable of interpreting unstructured workflows like invoices,
08:50like emails, contracts, and that enables automation across AR AP. Even if the digital payment,
08:57the payment itself is digitized, there's just so much manual intervention on both ends of the transaction
09:03that it just doesn't make the puzzle piece seamless and flow. Payments and fintechs are shifting from
09:11basic digitization to full straight through automation flows. MasterCard assesses $80 trillion commercial
09:19flows as an opportunity of which $63 trillion is in invoiced payments. When we tie that into the
09:25findings of our survey, the fastest areas where it's reaping benefits is around software development and
09:31engineering. 66% of companies have talked about scaled or deployed use cases rather than just doing
09:38pilots at the moment. Specific to SMB fintechs, AI adoption across these fintechs are centered around
09:47four high impact workflows. First is APR automation, which we talked about in the panel today. Second is
09:53better risk scoring and underwriting. Access to capital is one of the biggest challenges and AI can help you
09:58resolve that. Third is efficient onboarding and customer support. And all of them are aimed at supporting
10:05back office workflows to lift margin and to expand attach rates for fintechs.
10:11Picking through some of them individually. So we're picking out some case studies to show what
10:16companies are doing in that regard. So with Intuit, their AI bets deserve a mention. Our team thinks that
10:23its Gen AI Assistant Intuit Assist should provide a pricing tailwind and increased adoption of live experts
10:30that they work on their network. And with an end-to-end solution for small medium businesses with QuickBooks,
10:35AP automation, with their HR integration, all of that data comes into one place and you generate predictive
10:41insights out of that. That's a very, very valuable asset for SMBs. Frees up a lot of time to do more
10:47productive things and also makes it very sticky for Intuit. In this era where fintechs will, this is
10:55one space where I think fintechs will actually gain an edge. AI leaders are going to see lesser loss rates
11:02and volatility and higher tax rates for working capital. So if you monitor those statistics, you'll
11:08see how fintechs keep gaining an edge over banks because banks don't have their data sorted at the
11:13moment. At least not the primary banking relationships that you would think that would be targeting that
11:19segment. Risk underwriting is going to be continuous and adaptive and all of the conversations that we
11:26had around different data sets getting incorporated in that decision, how much is your AP, how much is
11:31your AR, your POS data, accounting feeds, tax information, all of that will become currency to
11:37get better credit access and availability for SMBs. Platforms that integrate AI into merchant-facing
11:45workflows, they can widen their competitive modes. All of you, this is something to kind of think
11:50about. Onboarding and support are one of the areas where we are seeing very early-scaled AI benefits.
11:56PayPal reported 1.5 times higher average margin per merchant and two times more product activations
12:04when they start onboarding their merchants with a survey and get to hear what the merchants are looking
12:10for and get to know them better at the start. MasterCard leverages AI across several use cases,
12:18including decision intelligence, etc. But I think the most notable that I have noticed is on the
12:23doubling down on digital identity and trust, where it has made sizable acquisitions, and I talked about
12:29Recorded Future. They block billions of dollars of fraud annually through AI-driven solutions, and they've
12:36boosted fraud detection rates by 20 percent to as much as 300 percent. And lastly, Block is positioning
12:45Square as an AI-native operating system for small businesses, embedding intelligence directly into
12:51merchant workflows. They've launched an LLM-driven tool that automates customer messaging, marketing, etc.
13:01The one thing that I really liked was their voice-enabled system, which basically, when I place an order on
13:08phone, that message does not get received by a person. It goes directly to their order platform,
13:13and the accuracy of that model is so much more better than what we've ever seen in the past, all thanks to AI.
13:21To end with, what does it take to win this race? Our survey confirms that doing the hard things
13:29will matter the most. Data remains the foundation and primary sticking point for AI adoption across
13:36all spectrums of financial services. More than a quarter of the participants identified data quality
13:41as one of their most pressing challenge, reflecting the industry's reliance on accurate, well-governed
13:47data sets for model reliability and also regulatory confidence. Another 30 percent ranked data privacy and
13:53cybersecurity as their top concerns. So that's what we think data readiness and data strength will
14:01determine winners. And it's not about who has the best AI model, but who has their data sorted and
14:07organized. With that, I'm on time. Thank you so much. Please reach out to us if you have any questions or
14:12you want to access more of this research. We're available on the terminal. Thank you.
Be the first to comment