How to Discover the AI Initiatives of Fortune 500 Companies

Daniel Faggella

Daniel Faggella is Head of Research at Emerj. Called upon by the United Nations, World Bank, INTERPOL, and leading enterprises, Daniel is a globally sought-after expert on the competitive strategy implications of AI for business and government leaders.

Artificial Intelligence in Corporate Banking - Current Applications

There is always a fascination with what the biggest and most powerful companies in any given sector are doing. If you look in the world of eCommerce, everybody is ultimately referencing Amazon. In the world of banking, firms like JP Morgan are referred to and everyone’s interested in their newest hires in terms of the C-suite and their newest innovations in terms of technology. Analyzing bigger firms seems to have a level-setting effect.

This is for a few reasons. First, it shows what the current winners are up to. We can presume that if a firm is the biggest, the strongest, the most profitable, the fastest growing within its given sector, that they’re doing something right.

Secondly, it informs a company’s own strategy. If the biggest firms are going to start competing on factor “X,” you might then consider factor X or ask if you need to position against it and go with factor “Y,” which might be some other aspect of customer service, for example.

It’s pretty important to know what the big players are up to if you’re going to have to respond in the same marketplace. So knowing what these leading companies are doing is important, but it has to be taken with a grain of salt, particularly in the space of artificial intelligence.

When a large business announces something that they’re doing with AI, it is a veiled assumption that it is an accurate reflection of what’s actually happening behind the scenes at these firms.

In this article, we’ll explore how we here at Emerj analyze what large companies are up to and how we look at what they say with a grain of salt, and what other sources we use to make some assumptions about what else might be happening behind the scenes. These include:

  • Assessing the purpose of a press release
  • Assessing the cost and priority of a particular application area
  • Assessing the vendor landscape

We’ve been covering the AI applications of the world’s largest firms for years. Before we explore some of our best practices for determining the AI initiatives of global firms, here is a selection of our findings from across industries and sectors:

Previous Coverage of Fortune 500 and Top Global Firms


Global / Other

For Brands, AI is About Perception

It’s important to note that the companies that are adopting AI are large industry-leading businesses. What they say is obviously a game of perception, not because they’re bad people, but because they’re in business.

When it comes to perception management, there are things that a company will say or admit to, and things that the company will not say, or not admit to. After assessing thousands of use-cases within a large firm, here are the trends what brand are willing (or not willing) to communicate via press releases and announcements:

What Brands Will Tell You

  • Anything That Makes Them Look Smart: First and foremost, what they will tell you has to be something that makes them look smart, advanced, or morally good in some way. It has to convey some kind of advantage to the customer. That’s ultimately what a press release is about. Maybe it’s about making the competitor shake in their boots, but really they want to be able to garner the proper reception from the folks that are going to pay them the money.
  • Anything That Makes Them Appear to Be Better Serving Customers: Customer-facing applications are remarkably common in this respect. If you look at press releases in banking, there is an over-representation of things like chatbots. That’s the case almost across the board if you look in different sectors.
  • Anything to Not Seem Left Behind: If a competitor does a press release or an AI application, then business might feel pressured to do the same press release even if their version of the technology is horrible or they only have a tiny demo to show for it. If there’s a tremendous amount of froth in a given industry, a good portion of that froth is meaningless; it’s just people saying that they too are doing what everyone else is doing, and that creates its own little wave.

Eventually, that settles down to an actual level of capability, which is almost always lower than the purported amount of capability that initially comes out during the frothy period.

What big companies will tell you is what makes them look good. Generally, looking good means it’s customer-facing. It’s giving convenience to the customer and making the customer think that the business is smart, that it’s an advanced company, that it’s not falling behind. Press releases are not intended for deep insight into business operations; they’re intended for perception management.

What Brands Will Not Tell You

These companies will not tell you if they launched a project and it failed. You also won’t learn about anything that could potentially garner bad press.

For example, we interviewed a gaming company once on the AI in industry podcast that went into some great detail about the way they use behavior within their mobile gaming apps to determine who are going to be the people that are willing to pay for them. Some people can spend hundreds of dollars a month on in-app purchases for these games.

What happened was some people found that episode and put it up on a gaming forum. They started really ripping into the company about it and finding those things out. Our audience may have liked him talking about how his company’s AI worked, but his customers didn’t like it.

The point is that there are likely going to be some marketing and advertising applications that companies are not going to hold press releases about. For example, when banks start adapting computer vision to recognize your face, they’re unlikely to do a press release about it. They might only do so to legally defend themselves.

But when we did our recent vendor scorecard report on AI in banking, we really couldn’t find any evidence of a bank outside of China stating openly that they were leveraging that kind of technology. It could be the case that banks just actually aren’t leveraging that kind of technology, but if they were, it’s unlikely anyone would hear about it.

Similarly, if a company launches a project or a product that would put them at risk if people knew about it, they might not speak about it. For example, it’s hard for banks to admit to using cybersecurity software because a large bank might not want their competitors to know exactly what software is protecting their most important data. It might actually be better and safer for those companies to never mention who they’re using for security in anything.

As a result, if a company like Darktrace lists a customer on their website, they must have jumped through a lot of hoops to get that bank to agree to Darktrace putting their name on their website. Such an admission likely does very little for the company’s credibility with its customers (as cybersecurity isn’t seen to be something that directly impacts customers), and could potentially lead to risk if hackers know what software they’re up against.

Companies offering chat applications or other public-facing, visible applications are generally in a much easier position to gather testimonials and named case studies on their website than a cybersecurity firm.

Outside Factors to Examine

Size of the Problem or Opportunity

You might also want to ask about the overall cost of the problem or the priority. In other words, you might ask how large a portion of the business is the area that the application is supposed to work for?

For example, in eCommerce, what is the cost of the call center versus logistics and supply chain?

Generally speaking, we tend to find the largest vendor companies are often in the spaces that have an intersection of two factors. The most money within a sector is often going to be where many of the biggest vendors focus in terms of artificial intelligence applications.

In part, this is because the people running those vendors understand that this is a priority because of how much it costs buyer companies every month. It’s also just that there are more people working there.

If it’s a bigger cost, there’s more overhead in terms of people, which means there are more people to potentially sell to. In the banking space, you will see a lot fewer press releases about compliance and fraud detection applications. But in fact, fraud is where most of the money is. So look at what the big costs are and try to see if there are any giant vendors in that space.

In addition, you can ask: “What are the problems that keep leaders in my sector up at night?”

In banking, we could summarize it as a risk. There’s probably a set of words that in general tend to be the C-suite priorities for all different sectors.

If you understand those C-suite priorities, it’s likely that there’s more potential AI acquisition in those space then maybe is being let on by the companies because again, there’s a strong bias for customer-facing applications.

Assessing the Vendor Landscape – A Critical Industry Assessment

This brings us to the last factor in analyzing what large companies are doing when it comes to artificial intelligence applications. And this is assessing the vendor landscape.

Now, admittedly, some of this could be challenging. This is what we do for our own clients (see Custom Research Services) when we’re hired to analyze a space; we’re almost always doing a vendor landscape assessment. When we look at vendors, what we’ll often find is the following four clues about where real AI adoption and traction is happening within a given space.

  • Customers They Claim: You can look at what a vendor claims as its customers. For example, we can assume that if an AI vendor selling into pharma lists Merck on their website that they actually have worked with Merck because otherwise, they open themselves up to legal repercussions.
    • The AI-based fraud detection software vendors we’ve seen selling to banks might list JP Morgan or Wells Fargo, but those banks don’t have any press releases of their worth with these vendors. So sometimes you’re going to have to learn about it on a vendor website.
  • Case Studies with Named Clients: You can also look at case studies on vendor websites. That said, case studies don’t count unless they name the client. If it says a global bank used a vendor’s solution, it shouldn’t be trusted. It’s also important to look at what percentage of the case studies that the vendor lists has quantifiable results.

In doing so, if we’re looking at something like payment fraud applications, we can get a pretty good sense of not only which of the big retailers in the world are actually using these solutions, but which of these solutions may have real evidence of results versus those that don’t?

You could do this across the entire eCommerce space of whatever sector you’re in, and then you’d be able to get an idea of what the norm is when it comes to AI adoption within your sector.

We do this in great depth across several strata with specific scoring methods that we use to actually look at the vendor landscape. But if you’re just trying to figure out where your biggest competitors stand when it comes to a specific AI application, you could read what they tell you, but you’ll want to take that with a grain of salt.

You can instead assess the vendor landscape and look at the factors I mentioned in this article.

Outside Fact

Take all communication from large Fortune 500 firms with a grain of salt, and understand that the perception they want to manage probably does not involve total transparency about their AI initiatives (who can blame them?). Knowing what PR communications are overrepresented or underrepresented can help business leaders understand what’s truly happening under the hood.

Here are four rules of thumb when assessing the AI use-cases of any large firm:

  • Anything customer facing that isn’t active is hype, or it failed
  • For every customer-facing AI press release, count half of them as hype
  • Assessing vendor case studies and claimed clients is a more accurate assessment of who is buying what than any press releases
  • Look at the 5 biggest vendor companies in a given sector (by funds raised or revenue). Whatever problems THEY solve are where much of the money is, and that’s often a better indicator of the Fortune500 in a sector than they press releases of the Fortune500 themselves.


Header Image Credit: Baseak

Stay Ahead of the AI Curve

Discover the critical AI trends and applications that separate winners from losers in the future of business.

Sign up for the 'AI Advantage' newsletter:

Stay Ahead of the Machine Learning Curve

Join over 20,000 AI-focused business leaders and receive our latest AI research and trends delivered weekly.

Thanks for subscribing to the Emerj "AI Advantage" newsletter, check your email inbox for confirmation.