What CIOs Say Is In The Way Of Successful AI Implementation

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.

---

As AI becomes less of a far-off technological idea and is more of an actual platform companies can readily implement, CIOs are making a potentially unexpected discovery: They’re not ready for it.

According to a new study from AI platform optimization provider Riverbed, 94% of C-suites make AI a key focus, with adoption being led from the top. But just 37% say they are actually ready to implement AI now. Organizations say they will be ready in the near future—86% expect to be fully prepared to implement AI projects in three years.

The big issue standing in the way of successful enterprise AI platforms is likely data. Nearly one in four AI projects are underperforming company expectations, and Riverbed suspects that is a data issue. After all, 85% say that good data is vital to quality AI, but about half of that number said a lack of high-quality internal data will keep them from investing more in AI. Companies’ view of their own data is especially low in the financial services industry, where only 36% said their data is excellent when it comes to quality and completeness, and 34% feel that way about its accuracy and integrity.

These results show it’s time for IT departments to build data readiness into their AI plans. Data needs to be accurate, available, and in a format that yields insights. Companies are loath to use sample data in order to build a working system, with 86% wanting to concentrate on their actual data. Riverbed found that high performers in AI—those already successfully using the technology for tasks like improving the digital employee experience—are much more likely to have their data in working order.

NOTABLE EARNINGS

Tech company Oracle saw its earnings surge after reporting blockbuster earnings this week, and its founder Larry Ellison’s fortune alongside it. The tech company not only delivered better-than-expected financials, with $13.3 billion in revenue—including 21% year-over-year increases in cloud revenue and a 45% increase in cloud infrastructure revenue—but also announced a new partnership with Amazon Web Services allowing customers to use Oracle’s AI applications within Amazon’s cloud. The surge increased Oracle’s market cap to from $386 billion to $435 billion, and boosted Ellison’s fortune $18 billion, elevating him to the world’s third wealthiest person.

Analysts expect Oracle’s growth to continue in the next couple of years, as its current strength is in data centers that large companies including AWS, Microsoft and Google are using to train their AI LLMs, writes Forbes senior contributor Peter Cohan, referencing commentary in the Wall Street Journal. On the earnings call, Oracle said it’s designing a data center that uses more than a gigawatt of power and will rely on three modular nuclear reactors. But what happens after the AI LLM training is done? Right now, data centers are good business, Cohan writes, but they might not be the best long-term bet, since their usefulness is limited.

LEGAL ISSUES

Google is dealing with antitrust accusations on both sides of the Atlantic Ocean this week. In the U.S., the tech behemoth is currently on trial to determine whether it has an illegal monopoly in the online advertising business.

The lawsuit behind the case was filed in January 2023 by the U.S. Justice Department and attorneys general from eight states—Virginia, California, Colorado, Connecticut, New Jersey, New York, Rhode Island and Tennessee. They said the tech behemoth has too much control over digital advertising because it owns the tools used to buy, sell and display ads, and has manipulated both market prices and the services that ad buyers need to use. CNN reported Justice Department lawyer Julia Tarver Wood argued Google has a “trifecta of monopolies” to control digital advertising: its ad server business for publishers, advertising exchange AdX, and its advertiser ad network. Google responded its services have made it successful, but that isn’t illegal. It also has made the case that other big players—including Meta, Amazon and Microsoft—have similar digital ad tech suites.

This case will be decided over the course of several weeks. But in Europe, the EU Court of Justice upheld a ruling demanding a €2.4 billion ($2.6 billion) fine from Google for giving preference to its own shopping recommendations ahead of those from rivals, writes Forbes senior contributor Emma Woollacott. This fine was levied in 2017, and was at the time the largest ever assessed by the EU. Google appealed it to the EU’s General Court and then the Court of Justice.

Kelkoo, a European shopping search engine among the plaintiffs in the case, was pleased by the ruling. “This decision is a win for fair competition and consumer choice,” the company said in an X post.

NOTABLE NEWS

Apple owes €13 billion ($14.3 billion) in back taxes to the EU, the EU Court of Justice ruled this week. At the heart of the issue is whether Ireland’s government had given unfair tax advantages to Apple, writes Forbes’ Kelly Phillips Erb. In an arrangement that had been in place since 1991, Apple attributed nearly all of its European sales to its Ireland “head office,” instead of the countries in which the items were sold. Because of Ireland’s agreement with Apple over corporate taxes, Apple paid an effective corporate tax rate of 0.005% in 2014. The EU began investigating this deal in 2014, after which Ireland changed its tax laws and Apple changed its policies.

The ruling, which is the final judgment after appeals, deals with taxes from 2003 to 2014. In a statement, Apple told Forbes it was disappointed with the decision, but it was never trying to evade taxes. “This case has never been about how much tax we pay, but which government we are required to pay it to,” the statement reads. “We always pay all the taxes we owe wherever we operate and there has never been a special deal.”

Also this week, Apple announced its new line of iPhones that CEO Tim Cook were “designed from the ground up” for AI. The iPhone 16 will hit the market next week, with preorders opening on Friday. Although this is Apple’s official entry into the AI market, the company’s stock slumped during Monday’s launch event and has made a slow recovery.

DEEP DIVE

Cops Are Testing Apple Vision Pro For Use In Surveillance

When the Apple Vision Pro launched earlier this year, it was touted as a revolutionary new contribution to mixed reality headsets—an area that has seen very limited adoption. And while analysts say the Vision Pro truly delivers a second-to-none spatial computing experience, sales have been much slower than anticipated, likely due to the $3,500 price tag and limited uses for the headset.

Police departments in California have found potential uses for the headset, and are piloting their use in law enforcement. Forbes’ Thomas Brewster writes that the Orange County Sheriff’s Department is testing out the Vision Pro as a way to view surveillance feeds, access law enforcement databases and chase down suspected crimes. The department has a new unit called the Real-Time Operations Center, which pools departmental data, ranging from license plates to cell phone locations to criminal records. The data is typically visualized on dashboards across several screens in a central location, but the department is testing making a virtual version of that center accessible from anywhere by using the Apple Vision Pro.

If the initiative is successful, Orange County Sheriff’s Department CIO Dave Fontneau told Forbes it could eventually eliminate the need for physical real-time data centers. Fontneau said he was able to put on the headset and put himself in a working virtual environment, navigating a city map, moving assets, placing officers in vehicles and manipulating camera feeds. While the department says the current model of the device is too heavy for officers to use on the street, a lighter version that is under development could be used by cops on the street to quickly access information about where they are, including neighborhood crime rates or recent illicit activities.

The Los Angeles Police Foundation also plans to spend $20,000 on five Apple Vision Pros. The headsets would be used for “enhanced surveillance and monitoring,” featuring augmented reality overlays featuring critical data and surveillance feed visualizations.

Fontneau told Forbes the department has been working closely with Apple on several projects, from training on the Vision Pro to deploying Apple CarPlay as a replacement for computers in police cars.

FACTS + COMMENTS

After two weeks of falling stock prices, Nvidia CEO Jensen Huang spurred a bit of a turnaround by talking about the large demand for its AI chips at a Goldman Sachs conference on Wednesday.

7%: Increase in Nvidia stock by late Wednesday afternoon

24: Days this year that Nvidia’s stock has seen a gain or loss of at least 5% on the market

‘Everyone wants to be first and everyone wants to be most’: Huang said at the conference

STRATEGIES + ADVICE

In this day and age, cyberattacks are inevitable, but there are things companies can do to be ready for them. Here are four areas to concentrate on to bulk up your cyber resilience.

It’s important to consistently show your team members how much they are valued. Here are 10 ways to let them know you appreciate them.

QUIZ

A North Carolina man was arrested last week on charges of using AI to commit what federal authorities called “brazen fraud.” What is he accused of doing?

A. Creating chatbots based on popular cartoon characters—including Bluey, Spongebob Squarepants and Mickey Mouse—without securing rights

B. Creating songs and continuously streaming them using bot accounts to generate royalty payments

C. Running an official-looking website that charges users to pick Powerball numbers, but does not buy them tickets

D. Running a Cameo-like service selling AI-generated celebrity videos of personal messages

See if you got the answer right here.