How CIOs Can Maximize ROI On Enterprise AI Platforms

This is the published version of Forbes’ CIO newsletter, which offers the latest news for chief innovation officers and other technology-focused leaders. Click here to get it delivered to your inbox every Thursday.

Even though neither of the candidates on major presidential tickets are talking much about AI, it’s a huge area of concern for tech leaders. Nearly three in four tech leaders said the election will have a major impact on whether the U.S. tech sector will be able to stay ahead of global competitors, according to a poll published this week by EY US. Specifically, they said, it will impact cybersecurity and data protection, AI and machine learning, and user data and content oversight.

There are very few issues that Vice President Kamala Harris and former President Donald Trump are even remotely close to agreeing on, so this result from the poll isn’t a surprise. But it’s worth taking a look at what may happen in terms of tech policy if either candidate is elected.

As a part of President Joe Biden’s administration, Harris supported the executive order he signed last year to place guardrails on the use and development of AI. The executive order mandates guidelines and standards for government use of AI, requires government oversight of AI models that could pose a risk to national security, and essentially sets the stage to ensure privacy, equity and civil rights are protected through the use of AI technology. Harris spoke about the order days after it was signed at the Global Summit on AI Safety in London, saying it is possible to ensure ordinary people are protected from AI misuse—like explicit deepfake images of them being shared, or insurance claims being denied because of faulty AI algorithms—and not stifle broader innovation. When Biden signed the executive order last year, several tech CEOs were immediately on board and had been working with the federal government to develop AI regulations.

Trump said last December he plans to cancel Biden’s executive order if elected, and that he’d also “ban the use of AI to censor the speech of American citizens on day one.” There hasn’t been much else published about what Trump would do instead, though a rundown by law firm Winston + Strawn indicates he has said that AI’s power to distort can be “scary,” and the threat AI could pose to security requires immediate attention. When Trump was president in 2019, he signed an executive order for the federal government to direct more funding to AI R&D and work toward building governance standards. Tech executive and self proclaimed “free speech absolutist” Elon Musk has been close to the Trump campaign, and Musk has used his social platform X to share his view that the First Amendment can protect campaign-related AI deepfakes and reproductions of copyrighted images. Trump has not endorsed Musk’s views, but he could be influenced by them.

But no matter who wins the election, 87% of tech leaders told EY they plan to increase AI investments by 50% or more in the next year, with most saying they want to reallocate existing resources toward AI. So even if the election has a huge impact on larger policy, businesses will stay on the journey to implement the technology to make their work more efficient and useful.

Even for a technology like AI, which many are committed to investing in, companies need to examine the ROI for systems they want to adopt. I talked to Anurag Malik, president and CTO of legal AI platform company ContractPodAi, about how to determine this and get the greatest value out of AI investments. An excerpt from our conversation is later in this newsletter.

Nominations for Forbes’ CIO Next List are now open. If you know any CIOs who are making an outsized impact on their company or business, please let us know about them here by November 8.

VALUATIONS

Apple’s stock surged to an all-time high on Tuesday, with the tech company briefly being worth $3.61 trillion. The likely catalyst was the announcement of a new iPad Mini model, which is the first update to that model since 2021. The new iPad Mini, which is available for pre-orders now and goes on sale next week, has a powerful A17 processor that will work with the company’s Apple Intelligence features when it launches later this month, writes Forbes senior contributor David Phelan. It also has 128GB of storage in its basic configuration, which is twice that of earlier models. Meanwhile, Morgan Stanley analyst Erik Woodring said lead times for the iPhone 16, released last month, are “encouraging,” and the 56 million iPhones shipped last month was 10% more than expected, writes Forbes’ Derek Saul.

Apple’s share price moderated a bit on Wednesday, but it will be both rolling out Apple Intelligence and reporting earnings at the end of the month, so more big changes can be expected.

Nvidia’s stock has also been rallying for much of October, and is nipping at Apple’s heels as the current second most valuable company by market cap. Nvidia’s stock surge, fueled by optimism and encouraging words about demand from CEO Jensen Huang, briefly moderated on Tuesday. Dutch firm ASML, which makes equipment used to manufacture high-end AI chips, mistakenly reported earnings a day early and warned of “more gradual growth” to come, with 2025 revenue projections well below analysts’ expectations. ASML’s share price fell 17.5% on Tuesday, bringing down stock for Nvidia and other AI chip companies, including Advanced Micro Devices, Arm Holdings and Broadcom. But the rally picked back up again Thursday morning, when Taiwan Semiconductor Manufacturing Co. reported blockbuster earnings and increased its revenue guidance. TSMC’s stock went up more than 12%, boosting its market cap above $1 trillion, and taking chip stocks including Nvidia, AMD, Broadcom and Intel with it.

FUNDING

Two startups focused on improving AI platforms raised large Series B rounds this week. Decagon, a platform that builds AI customer service agents, raised $65 million in a round led by Bain Capital Ventures, with participation from tech investor Elad Gil and VC firms Accel and Bond Capital, writes Forbes’ Rashi Shrivastava. Decagon, whose agents are used by companies including Notion, Duolingo, Rippling, Bilt and Substack, is now valued at $650 million, with $100 million in total funding to date. The company plans to use the funds to grow its team, go after new types of customers and add a voice feature.

Galileo, a startup that evaluates AI models to ensure they aren’t hallucinating or giving out proprietary data when prompted, announced a $45 million funding round, writes Forbes’ Richard Nieva. The round was led by Scale Venture Partners, with participation by Premji Invest, venture arms of Citibank and the data storage company Databricks and Hugging Face founder Clem Delangue. The three-year-old company, which offers four products to help companies train and evaluate their AI systems, says it has raised a total of $68 million to date.

TECHNOLOGY + INNOVATION

As technology changes, competitors are becoming allies. Intel and AMD announced the creation of an x86 ecosystem advisory group this week, dedicated to bringing tech industry leaders together to come up with consistent and compatible implementations for a variety of applications and platforms using x86 architecture. Forbes senior contributor Patrick Moorhead talked to Intel CEO Pat Gelsinger and AMD CEO Lisa Su about the rival companies—they still will be competitors, Su and Gelsinger told Moorhead—coming together now to give x86 processors the same kind of alignment found in the Arm processor ecosystem. Su told Moorhead that customers have likely been wanting this kind of arrangement for decades to make it easier on their software stack, and working with Intel on the standards is “a major benefit for our customers and for the ecosystem.” The two CEOs said the industry cooperation will help speed innovation and build market opportunities for x86 processors going forward. Most of Big Tech is represented in the group, including Meta, Microsoft, Google, Broadcom, Dell, HP, HPE, Lenovo, Oracle, Red Hat, Tim Sweeney of Epic Games and Linus Torvalds of the Linux Foundation.

BITS + BYTES

ContractPodAi CTO Shares The Keys To Finding AI ROI

Whenever a company wants to add new tech, even the most bullish will question its ROI. I talked to Anurag Malik, president and CTO of legal AI platform company ContractPodAi, about how companies can get the most out of their investments in AI technology. This conversation has been edited for length, clarity and continuity.

How do most companies go about trying to figure out the ROI of generative AI?

Malik: We are all running a business. Most companies buying a product are doing it for return on investment, with or without AI. I think the onus lies on the product provider to show that value. Value can come in many ways. It could be time saved, people being redirected to something more efficient, cutting down on outside spend. We think of AI as something that’s not going to take jobs away, but it’s going to make you much more efficient. You’ll be able to do much more of what you used to do before with limited resources, especially repeatable tasks, mundane tasks, which is very easy for AI to pick up.

One thing we do is show productivity dashboards within the product, which shows time saved, number of tasks finished in whatever amount of time. We let our buyers input how much time it takes them today. Then we do a comparison: This is how long it’s taking you now, and you can put in a dollar value to the labor and then get ROI.

One of the shortfalls is people tend to map existing processes to a technology, which is very different from how they were do[ing] things manually before. Helping them bridge that and re-imagining how they would use the technology to do things different[ly] and be much more productive is something the provider should do because the buyer is new to our technology and it’s on us to guide them. It’s more of a consultative way of selling AI and educating the customers as we go along.

When a company is looking to adopt AI, or any sort of tech system that will definitely change the way they do things, should they be thinking ROI first, or should they be thinking about long-term capabilities?

Being long-sided on this is way more valuable. In fact, when we onboard customers, we tell them, “Hey, we all walk and run.” As cliche as it sounds, you can’t go from zero to 100 miles an hour in two seconds. This is not a Tesla. This is a new system. There’s people involved who have to learn to use it. So let’s get the low-hanging fruits. You may not see it’s better than what you do today. You may not see humongous ROI, but if you get people on board and make them get used to the product, the road down is going to look much more productive for you. Absolutely, that long-term vision is very necessary.

Software providers have to educate their customers to see that, to give them that vision. Because remember, everybody who comes to buy technology is trying to solve a pain point today. They’re not trying to see what’s going to happen in three years down the line. Yes, they’re invested in it, but if you educate them well enough that, hey, you’re not only going to solve this pain point today, but if you take a step back, you can see the vision of how you can really transform how you’re doing things.

What are some of the things that get in the way of companies realizing the ROI potential that they have with AI?

Adoption. The distance to change is the biggest thing I’ve seen. Especially global companies: different regions, different teams. For it to go from one region to another is the biggest challenge.

One thing I’ve seen is if you get the first one right and make a good story of it, educate teams wherever in the world they are, it works very well. The first success story is very important. It doesn’t matter how well your software works. If people are not going to use it, it’s as good as crap. Adoption and change management—making people understand [they need] to change the way they do things—is very important. Even more important than the actual technology itself.

FACTS + COMMENTS

Amazon entered into agreements with Dominion Energy, Energy Northwest and startup X-energy to fund development and deployment of small modular reactors—small and advanced nuclear reactors—to create the power needed for AI datacenters.

$500 million: Amount invested in X-energy for these reactors, anchoed by Amazon

960: Megawatts of power to be created by the four SMRs Amazon is helping Energy Northwest develop. This is enough to power the equivalent of 770,000 homes

‘Nuclear is a great option for how we scale the world’s energy needs’: Amazon Web Services CEO Matt Garman told Forbes

STRATEGIES + ADVICE

Executives are faced with an avalanche of tasks that can be difficult to prioritize. Here’s how to use the Eisenhower Matrix—a streamlined and simple way of thinking about things—to help decide where to invest your time and energy.

Inspirational leaders can set our professional journey, and you can be one, too. Here are four tips to set you on that path.

VIDEO

QUIZ

A foundational online resource that users have relied on for decades was hacked and taken offline for a short time last week. What is it?

A. Wikipedia

B. Wayback Machine

C. Tripadvisor

D. Craigslist

See if you got the answer right here.