2025 Is the Year of Scaling AI
11:18 Share

Hot off the presses, we have some extremely interesting statistics about how big companies are thinking about generative AI. So here is all of the latest in enterprise AI. The AI Daily Brief is a daily podcast and video about the most important news and discussions in AI. To join the conversation, follow the Discord link in our show notes. ♪

Hello, friends. Quick note, today's episode was recorded in advance of its release date, which means no headlines today. We will be back to our normal format tomorrow, but for now, enjoy these interesting findings. Welcome back to the AI Daily Brief. Today, we're doing something a little bit different. My friends over at KPMG gave me some early access to their quarterly pulse survey on AI. And there's some really interesting stats, a lot of which maybe validate the sensibilities that we've had, but finally put some real numbers around them.

To get us started, I always like to understand what went into this survey, who was consulted, what the sample of companies actually represents. These came from surveys of US-based companies with a billion dollars or more in revenue. Of the 100 people surveyed, 20% were the CEO or president, another 30% were at the C-level, and 50% were executive vice president or managing director level. It was about two-thirds, one-third public versus private companies, and 51% of respondents represented companies with 10,000 or more employees.

Those surveyed represented a pretty wide cross-section of functions, 20% in tech and IT, 17% in HR workforce and talent, 18% in finance, 14% in risk, 13% in marketing and sales, and 18% operations. So pretty wide cross-section of very large companies is the TLDR. What we're going to talk about today is we're going to break things down into agents, attitudes, leadership and phasing, investment sizing, challenges, and ROI.

And let's not bury the lead. AI agents are very much a big theme. Some of their highlighted stats, approximately 51% of organizations are exploring the use of AI agents with another 37% piloting. That means the vast majority are now in the agent space in some way or another. At the same time, the number of organizations that have actually deployed AI agents is extremely low. It's about a tenth of the companies surveyed, just 12%.

Maybe more interesting is the details they found around what use cases and functions people were looking to agents for.

At this stage, the most prominent use case appears to be analyzing complex datasets, with 70% saying they already use an AI agent for this, and only 7% saying that they didn't and they don't have any plans to. 54% say that they have plans to integrate call center agents, with another 16% saying they're already doing that. And 60% say they have plans to have agents perform administrative tasks like scheduling meetings, with 27% saying that that's already happening as well.

One interesting use case that seems kind of divisive, 27% said that they already have an AI agent conducting employer reviews, but only another 30% said that they have plans to do so. That compares to 43% said that they don't have plans. So this seems to be something that people are very interested in or very not interested in. The big takeaway here is that we are very much heading into the pilot phase of AI agents. Just a tiny percentage are actually deploying, but everyone is interested.

Next up, I wanted to talk about attitudes. One of the things that has been very notable about generative AI is the level of optimism and excitement leadership has around it. Certainly that extends to this survey as well. In the next year, 70% of leaders who are surveyed say that they believe that AI is going to help their organization run a better business. A full 56% said they believe that AI will fundamentally change their organization's business. Those numbers jumped to 91% and 67% respectively if you zoom out two years instead.

Interestingly, when it comes to the why of AI, when KPMG asked to what extent are the following factors influencing your AI strategy today, the number one response was macroeconomic factors like GDP growth and inflation. Still, this is kind of a story of an all-and kind of answer, as every single factor that they checked on had at a minimum about half of people say that that was influencing their strategy. That includes everything from the opportunity to identify new revenue to pressures to improve efficiency.

Today's episode is brought to you by Vanta. Whether you're starting or scaling your company's security program, demonstrating top-notch security practices and establishing trust is more important than ever.

Vanta automates compliance for ISO 27001, SOC 2, GDPR, and leading AI frameworks like ISO 42001 and NIST AI risk management framework, saving you time and money while helping you build customer trust. Plus, you can streamline security reviews by automating questionnaires and demonstrating your security posture with a customer-facing trust center all powered by Vanta AI.

Over 8,000 global companies like Langchain, Leela AI, and Factory AI use Vanta to demonstrate AI trust and prove security in real time. Learn more at vanta.com slash nlw. That's vanta.com slash nlw.

If there is one thing that's clear about AI in 2025, it's that the agents are coming. Vertical agents by industry, horizontal agent platforms, agents per function. If you are running a large enterprise, you will be experimenting with agents next year. And given how new this is, all of us are going to be back in pilot mode.

That's why Superintelligent is offering a new product for the beginning of this year. It's an agent readiness and opportunity audit. Over the course of a couple quick weeks, we dig in with your team to understand what type of agents make sense for you to test, what type of infrastructure support you need to be ready, and to ultimately come away with a set of actionable recommendations that get you prepared to figure out how agents can transform your business.

If you are interested in the agent readiness and opportunity audit, reach out directly to me, nlw at bsuper.ai. Put the word agent in the subject line so I know what you're talking about. And let's have you be a leader in the most dynamic part of the AI market.

Now, when it comes to where we are in the sequence of AI, obviously, we got a little bit of that with agents. But more broadly speaking, if we're looking at generative AI as a whole, there is definitely progress being made from a pilot stage to actually getting to scale.

I think this is best summed up in the question where they asked what best describes the phase your organization is in in its Gen AI journey. Six months ago, 55% of respondents said it was best summed up as research and development or understanding the technology and its potential. 14% six months ago said that they were experimenting with proof of concepts or pilots.

Another 20% said that they were in strategic planning, such as establishing a roadmap, KPIs, and data management. Only 10% six months ago said that they were scaling. That number today has jumped all the way up to 50% who say that they're scaling the technology, which KPMG identified as enterprise-wide adoption and optimization and measurable outputs. Research and development, which was 55% six months ago, is now all the way down at just 8% today.

Interestingly, no one said that they were at established ROI phase now, but 31% said that they believe that they'll be at established ROI phase six months from now. So clearly a lot of optimism that organizations have this tiger by the tail and are ready to move. Now, when it comes to how much businesses are investing, one of the headliner statistics was

was this one. 68% of these leaders are going to invest between 50 and 250 million in Gen AI over the next year. That's up from 45% in Q1 a year ago. Basically, the scale of spend is increasing significantly. And remember, that's inside a single year because these are Q4 2024 statistics, not Q1 2025 statistics.

Overall, the most common investment size anticipated over the next 12 months was $50 million to $99 million, with 49% of the respondents placing their budget in that range.

Now let's talk about what challenges organizations see on the horizon. Data is by far the big leader. 85% said that they expected the quality of organizational data to be an issue with their Gen AI strategy in this year. That's a whole lot of organizations hoping the models advance so that we don't really have to care about data readiness anymore.

71% of respondents pointed to risk management, such as data privacy and cybersecurity as their big issue, and nearly half, 46%, pointed to employee adoption. Obviously, employee adoption is something we spend a lot of time on here at Super, and it's definitely a real challenge. Interestingly, there's a whole different set of questions around risks when it comes to agents.

For example, 29% indicated that they're not yet comfortable with autonomous agents and will require human-in-the-loop oversight. 11% said that they're developing AI agents only in-house. 31% said that they're not allowing AI agents to access sensitive data without human oversight. And 47% said that they're looking at AI agents as augmented support for their employees. Basically, the idea here is that even if agents could replace entire people, that doesn't seem to be, at least right now, the way that organizations are looking at it.

Lastly, let's talk about ROI. I mentioned before that no organization said that they're in the ROI phase right now. However, 31%, nearly a third, anticipate that that's where they'll be in six months from now. And frankly, as much as we talk about ROI, it's very clear that we are still operating in this setting where there is such a strong presumption that AI is going to be so transformative that no one is stopping or slowing down because they haven't exactly figured out ROI yet.

In terms of how people are thinking about ROI, there has been a big resurgence in a focus on productivity. Only 36% of organizations said that they're measuring ROI based on increased productivity in Q3, but that jumped all the way to 79% in Q4. Profitability is the second most common measure of ROI at 73%. Revenue sits at 41%. And a set of other ROI metrics like employee adoption, employee AI learning and development are all somewhere between 10% and 15%.

One of the things that I'm watching most closely is how the introduction of AI agents changes the nature of the ROI discussion.

I can kind of see it going both ways. AI agents implicitly have a clearer path to ROI in the sense that if they work, working necessarily means doing labor that humans do right now for cheaper. Now, that, of course, does not mean that organizations are going to fire everyone. How they choose to redeploy those savings in terms of time and capital is going to be an organization by organization decision.

I've made it pretty clear on this show that I think that the organizations who win AI, quote unquote, will be those who choose to reinvest their gains from things like agents in better services and new products. But at the same time, there's no denying that the ROI is clearer from agents, at least theoretically, than it is from co-pilots and the current crop of human assistant AI. So will the introduction of agents put pressure on ROI metrics? Because since we are able to measure things with AI agents, shouldn't we be able to everywhere else as well?

Or will there be such a focus on agents that it'll take some ROI pressure off the other areas? I think that'll be an interesting thing to watch in the coming months. For now, really interesting stuff from this pulse survey. Quarterly changes, especially now with things moving so fast, are a great vector for understanding. So thank you to KPMG for sharing this early and appreciate, of course, all of you guys for listening or watching. Until next time, peace.