How AI will transform business in 2025

2025 was the year that artificial intelligence (AI) strengthened its position in global business.

Throughout the year, OPTO session We spoke to leaders from a variety of sectors, including lending, customer service, and cloud storage. Here's what they have to say about how AI is impacting business and reshaping global markets.

Viable use case

Fully autonomous customer service agents. 24/7 Mortgage Advisor. A smart voice assistant built into your car's dashboard. Fast loans approved by AI.

These are just some of the innovative AI use cases talked about by interviewees during 2025.

What specific benefits can consumers and businesses expect from AI?

While much of the talk about AI focuses on improving efficiency for businesses, advances in agent AI could lead to real wins for customers, said Leah Price, head of the Tinman AI platform at Better Home & Finance. [BETR]I will explain. “If I'm a borrower and I work a 9-to-5 job, I can't use the user interface…Soon, you'll be able to talk to one of our avatars in the middle of the night. You're nervous about applying for a mortgage. You want to FaceTime with a professional mortgage advisor. It's possible.”

These benefits extend to client companies as well. Des Traynor, co-founder and chief strategy officer at Intercom, envisions Intercom's Fin AI to become an agent that can learn and adapt to internal changes within an enterprise. “Our goal with Fin is to have a human support team answer things first and last.”

To Paul Gu, co-founder of AI lending company Upstart [UPST]the victory is in the time saved. “Why do you do that? [people] Do you care about credit? That's because they want to become homeowners someday. They want to be able to buy a home when they need it. And they're probably hoping to be able to get a loan quickly enough to actually go under contract on the home they're buying without running out of credit. [of time] Or someone might occupy your home. ”

When you add voice as an interface to AI, the use cases are endless. SoundHound's [SOUN] Nitesh Sharan, Chief Financial Officer (CFO), testifies. “Our vision is to voice-enable the world with conversational intelligence, so that doesn't mean it's specific to any particular industry…We're trying to enable transactions, commerce, and discovery through all the ways consumers want to discover.”

bill payment

It's easy enough to judge the potential uses of AI, but some skeptics may wonder where the money will be made in the long run.

Currently, chip makers like Nvidia [NVDA] And applications like OpenAI get most of the media coverage. However, the box [BOX] CEO Aaron Levie argues that the value of AI is starting to trickle down to a broader range of businesses. “At the application layer, we're finally at a point where companies can leverage AI across their business processes. I think you'll see that value flow into the software layer.”

As many onlookers worry about the possibility of an AI bubble, it's helpful to consider how continued advances will increase the value AI brings to businesses.

When it comes to customer service, the focus is simple and “proactive support,” says Intercom's Traynor. “It's about how do you proactively resolve user issues, and how do you detect users who are having problems, who are misconfigured, who are about to have a problem? Every time you release a product and there's a flashing red light or an error message, you really have to assume that's the start of a support conversation.”

Upstart's Gu is optimistic about his company's role in applying AI to consumer finance. “In five years, I think we'll be successful in bringing AI lending to really big, important parts of the consumer credit market that matter to people: auto, home, revolving credit,” he says.

“When that happens, I think there will be a big shift in people's perception of the relevance of AI and lending today.”

And that could mean fairer access to credit for more people. “The end result will actually be asymptotically approaching the true level of access to credit that should exist in this country. Something like 80 percent of Americans should have access to high-quality bank credit. They should be able to use it on demand and with virtually no work required. And the remaining 20 percent should have a clear, defined path to get from where they are to where they want to be.”

Box's Levie emphasizes that behavioral change can be an even bigger challenge than technological advances. “This will take years. Change management in large enterprises is not as simple as getting people to use a new set of tools with new behaviors.”

New tools, but also new behaviors, are some of the things that make the AI ​​revolution so revolutionary, he says. “For a lot of software over the last 10 to 20 years, there hasn't really been a behavior change; just the tooling has changed,” he explains. Regarding AI, “What we're getting here is a change in behavior. The way we actually interact with the software is completely different. I'm no longer doing all the work myself within the software. I'm actually interacting with an AI worker on the other side.”

organize the slop

The fact that Merriam-Webster Dictionary selected “AI slop” as the word of the year is telling. This cutting-edge technology will really take off in 2025, but the hype surrounding it and the billions of dollars invested in AI-centric computing and infrastructure support make it difficult to tell what's working and what's not.

Leah Price's background in the public sector gives her experience in determining the true value of AI solutions. “Essentially any company can go out and claim, oh, we have all this AI, our efficiency has increased so much, it’s so cool, they flaunt the chatbot they slapped on top. And there you have it, they’re an AI company. They need to trade at higher multiples and that way they can get a lot of attention.”

So how should investors know where the true AI innovators are?

“You need to do your own homework.”

Although it's easy to be moved by numbers, Mr. Price insists, “Don't just listen to financial results announcements.'' Instead, try out the AI ​​solutions that companies are promoting to see if they really work. “Push our AI hard. Ask her the hard questions,” she says of Better's Betsy AI chatbot. “Then try visiting other companies' websites. Compare them side by side. See how far you can go.”

Ultimately, Price explains, more discerning consumers will require higher-performing AI solutions.

“If I can make my competitors feel even a little uncomfortable, that's when I know I've done my job,” she says.

Intercom's Traynor agrees. “AI is rife with what I call obsolescence,” he explains. “It's something I bought, it's been on the shelf, and no one has touched it in a year.” An investor himself, he outlined four criteria for determining whether a so-called AI-first company is genuine.

First, “It's very important that the product is used by customers.”

Second, it has to solve a real business problem, not what he calls a “cool tool.”

Third: “We need highly differentiated AI…It’s not something you can build in a weekend by just handing off the whole problem to OpenAI or Anthropic.”

Finally, and most importantly, “we just need evidence that margins can turn positive… It's very important that we don't spend staggering amounts of tokens doing something that has no value at all.”

At the end of the day, it all comes down to good business. “If you're burning an entire haystack to find a needle, and people don't really care about the needle, that's not an efficient business.”