Artificial intelligence can lead to surprises in all sorts of places where a bill once would have been considered settled.
The use of AI by Hertz (and European car rental company Sixt) to scan for damage on cars, which is then charged to the customer, is a new application of the technology that is creeping into consumer life unnoticed. But it won’t be the last unexpected adjustment to the travel experience courtesy of AI.
Experts say consumers should expect to see businesses across the service industry deploying similar technology in the future, if they aren’t already.
“As businesses seek to automate loss prevention and operational efficiency, we’re witnessing the emergence of what I call ‘algorithmic auditing’ – the systematic deployment of AI to identify, classify, and monetize previously overlooked inefficiencies or losses,” said Shannon McKeen, professor of the practice and executive director for the Center for Analytics Impact at Wake Forest University School of Business. The Hertz program, recently reported on by the New York Times, is the beginning of what McKeen describes as a broader transformation, and new fault line, in the service economy.
“The implementation of these systems reveals a fundamental tension between operational efficiency and customer satisfaction and equity,” McKeen said. The question isn’t simply whether AI can detect a scratch on a rental car bumper. “It’s whether businesses should charge customers for every microscopic imperfection that algorithms can identify but human judgment might reasonably overlook as normal wear and tear,” he said.
McKeen says the dialogue between service agent and customer over costs will increasingly include a new term: “the machine says.”
Scanning the hotel room
Hotels are working their way through these changes, according to Jordan Hollander, cofounder of Hoteltechreport.com, a research platform that helps hotels find new digital and AI products to improve efficiency.
“I’ve been seeing more hotels experiment with AI across operations, but not quite in the same way Hertz is using it for automated damage detection and billing. That said, we’re not far off,” Hollander said.
Some hotels, for instance, are already using AI-powered sensors to monitor air quality and trigger fines for smoking or vaping in rooms. But Hollander warns that sometimes the sensors trigger false positives.
“Like someone using a hairdryer or aerosol spray — and guests get hit with $500 charges without ever lighting up. It’s not hard to imagine how that could go south quickly,” Hollander said.
But unlike the car rental example, most hotels haven’t automated the billing step yet.
“They’re using AI more to flag potential issues — like a room that smells off, linens that don’t meet standards, or maintenance problems — and then looping in a human for the final call,” Hollander said. For now, the AI is acting more like a very observant assistant than a judge and jury.
“But it’s clear that hotels are heading in the same direction,” he said. “Between computer vision that can detect damage or wear in a room, and AI that analyzes guest behavior or room conditions in real time, the tech is already there.”
Risk of customer backlash
In a hospitality industry where trust is everything, there are reasons for hotels to move with caution. To date, many hotel operators are using AI to improve things like housekeeping efficiency, energy usage, and guest messaging — but they’re being cautious about when and how it impacts the guest directly in a way that can be perceived to hurt the experience.
“There’s a risk of backlash if hotels start billing guests based solely on what an algorithm says. The moment a guest gets a charge and can’t get a straight answer about why or how it was verified, you’re in dangerous territory,” Hollander said. “If guests feel like they’re being watched or nickel-and-dimed by a machine, it undermines the relationship completely,” he added.
Recent experience in the hotel industry provides at least one cautionary tale, according to Hollander, referring to a custom-modified Alexa for hotels. “Years ago, the hot thing was voice devices, and that never really took off for this reason,” he said.
A Hertz spokeswoman told CNBC that AI brings uniformity and consistency to the checkout process.
“For years, vehicle damage inspections have caused confusion and frustration. The process was manual, subjective, and inconsistent, and that isn’t good enough for our customers or our business,” she said.
She added that with digital vehicle inspections, Hertz is introducing “much-needed precision, objectivity, and transparency to the process – giving our customers greater confidence that they won’t be charged for damage that didn’t occur during their rental, and a more efficient resolution process when damage does occur.”
Of the 500,000 rentals scanned so far, more than 97% showed no billable damage, according to Hertz, and damage incidents are declining at scanner-equipped locations.
The Hertz spokeswoman acknowledged that the new system is still a work in progress.
“We know change of this scale takes time, and we’re listening, learning, and improving every day. As we said from the start, our goal through this initiative is to enhance the safety, quality, and reliability of our fleet and to create a more consistent rental experience for our customers.”
AI excels at pattern recognition, but where it may fall short is with the nuanced decision-making that has historically characterized good customer service, according to McKeen.
“What makes these systems particularly problematic is the erosion of contextual judgment,” McKeen said. Traditionally, business relationships relied on human discretion to navigate gray areas like “when does a scuffed tire represent normal use versus chargeable damage? When does a hearty portion in a restaurant satisfy a hungry customer versus being wasteful?”
Other companies will be watching Hertz closely to see how the AI experiment works out, he said, and then jump right in on the profit opportunity if it is determined that use of the technology won’t drive customers away.
Automation vs. ‘absolute overkill’
The use of AI for cost recouping isn’t widespread yet because companies have not figured out the balance between customer trust and implementing AI, and the benefit, so far, doesn’t outweigh the potential loss in loyalty, said Chuck Reynolds, managing director at L.E.K Consulting and a member of the firm’s digital practice.
The key for companies to implement these cost recouping tools is transparency. “While the opportunity for AI is huge, organizations need to be thoughtful about embedding it as a copilot, not police or enforcer,” Reynolds said. Sustomers will accept AI as part of the experience, he added, if companies are fair, visible, and design the AI experience with empathy.
“AI has to have customer-centricity built into its core,” Reynolds said, and companies have to keep a role for humans in the process to oversee and override the AI if necessary. “Organizations that do so without thinking through the entire process will have challenges with internal adoption and customer adoption,” Reynolds said.
Customers should expect to see more of the technology Hertz is deploying in different settings, according to David Rivera, professor of hospitality and tourism at Flagler College. In addition to hotels, the future could include restaurants using AI to itemize plates to ensure accurate billing. But Rivera says all of this is being done with the goal of operational efficiency rather than to punish the customer. Use of AI in hospitality is evolving from passive data collection to active use of real-time decision-making tools, Rivera said, and that includes things like monitoring your rental car or how much you are raiding the mini-bar in your hotel room.
“The common thread is increased operational efficiency, enhanced guest satisfaction, and automation of traditionally manual tasks, with a layer of accountability and transparency for both guest and provider,” Rivera said.
Not everyone is on board with that view, however.
“This trend is absolute overkill with AI solution capabilities,” said Daniel Keller, CEO of cloud infrastructure company Influx Technologies, which provides data collection and data analysis tools. “This particular use of AI doesn’t increase efficiency; it scrutinizes customers of small-margin service businesses looking to suck extra money out of guest experiences.”
Delta is pushing back against allegations that it is using AI to price discriminate.
The airline’s chief external affairs officer, Peter Carter, penned a letter on July 31 responding to a series of detailed questions put forward by Senate Democrats Mark Warner of Virgina, Richard Blumenthal of Connecticut and Ruben Gallego of Arizona.
“There is no fare product Delta has ever used, is testing or plans to use that targets customers with individualized prices based on personal data,” Carter wrote.
The controversy stems from Delta’s partnership with Fetcherr, a company that deploys AI to generate real-time dynamic pricing. In a July earnings call, Delta president Glen Hauenstein said that Fetcherr-assisted fares made up 3% of its domestic network and that the airline’s goal is to raise that to 20% by the end of the year.
At Delta’s Investor Day last November, Hauenstein spoke at greater length about Fetcherr, calling it “a full re-engineering of how we price and how we’ll be pricing in the future.”
He explained that airlines traditionally set price points, then use revenue management to control access to the inventory of those price points. But over time, those functions will be melded together into a single process of offer management.
The airline, he added, “will have a price that’s available on that flight, on that time to you, the individual.” Early testing, Hauenstein said at that time, had shown “amazingly favorable” revenue versus traditional pricing.
Now Delta’s plan to rapidly accelerate its usage of AI-assisted pricing, coupled with Hauenstein’s reference to individualized pricing, has generated a backlash.
On the American Airlines earnings call on July 24, CEO Robert Isom said American will deploy AI only for matters that help travelers, such as operations, product display and employee efficiency.
“This is not about bait-and-switch,” Isom said. “This is not about tricking, and others that talk about using AI in that way, I don’t think it’s appropriate. And certainly, from American, it’s not something we will do.”
Meanwhile, in their July 21 letter, Warner, Blumenthal and Gallego asked Delta to explain which data it is using to set prices.
“Delta’s current and planned individualized pricing practices not only present data privacy concerns, but will likely also mean fare increases up to each individual consumer’s personal pain point at a time when American families are already struggling with rising costs,” the letter reads.
In his response, Carter said Delta has no tolerance for discriminatory pricing. Delta’s AI pricing engine uses aggregated data, he said, including purchasing and demand data for specific routes and flights. The tool can also help Delta’s data analysts adapt to new market conditions and can factor in thousands of variables simultaneously.
The pricing tool, Carter added, recommends pricing adjustments both upward and downward, “benefitting both our customers and our business.”
Gallego, though, isn’t satisfied. In a statement issued Aug. 1, the Arizona senator said that Delta is telling investors one thing and the public another. He called for further clarification on whether Hauenstein misspoke during the airline’s Investor Day.
NEW YORK — Expedia Group, an early adopter of AI, has deployed the technology throughout its platform and reports tangible results. Development executives discussed it during the Explore Local event held here.
Take, for instance, Scout, an AI-powered system the company created. Scout makes recommendations to Expedia’s hotel partners based on what similar properties in the marketplace are doing.
“Based on these recommendations, our hotel partners last year took a million actions with great results,” said Greg Schulze, Expedia’s chief commercial officer. “These actions resulted in 10% incremental transactions and $6 billion of incremental revenue to our partners.”
AI is also “turbo-charging” Expedia’s advertising products, Schulze said, with artificial creative intelligence that generates images and machine learning-powered programs that select the right images for travelers.
Hari Nair, senior vice president and general manager of Hotels.com, provided an example of how Hotels.com is using AI to improve the traveler’s experience. The OTA was among the first to introduce AI filters. Travelers can search for “a hotel with a balcony and a sauna,” he said. The top three most-searched terms are hotels with parking, free breakfast and all-inclusives.
“The conversion in this case is almost 1.3 times the average conversion rate that we see on our site,” Nair said of users who use AI filters.
On Expedia’s consumer-facing app, a new capability launched in January that utilizes AI to analyze more than 2 million flights a day, said Tracey Weber, senior vice president and general manager of the Expedia brand. Its aim is to find fares “that are at least 20% better than the predicted price,” then alert consumers.
“For our travelers, this is a first step in figuring out where might be a great place to go and really inspire them as a demand driver,” she said.
It’s also a conversion driver, according to Weber, with 15% higher conversion with travelers who use the feature. Additionally, an AI itinerary builder has gotten a good reception among consumers. Weber said 65% or more of Expedia’s customers who use the feature describe it as “a delightful experience.”
The importance of AI at Expedia
In a press conference, Schulze said AI is “in our DNA.” Sam Altman, CEO of ChatGPT creator OpenAI, previously sat on Expedia’s board. Today, Alexandr Wang, founder of Scale AI and Meta’s AI chief, is a board member.
Expedia Group, Schulze said, thinks of AI in three main buckets. First is improving Expedia’s experience for travelers and partners alike. Second is working with AI-native companies like OpenAI on several fronts, ensuring travelers are still being directed to Expedia and that the company is providing the right kind of content — and accurate content — for AI engines to consume. Third, he said, is productivity.
The Scout example he gave on stage is an example of the third bucket, Schulze said. In addition to providing actionable recommendations to hotel partners, Scout is also deployed internally with Expedia’s sales and technology teams.
Another internal feature Expedia has created for employees is a framework that helps them choose which Large Language Model (AI models behind generative AI platforms like ChatGPT) is right for any given purpose, according to Karen Bolda, chief product and technology officer of Expedia’s B2B business.
“There are over 60-plus models there, because certain models are better at certain things,” Bolda said.
When it comes to AI, Expedia is always considering whether it should partner with other companies that have developed specific products or develop its own in-house, Bolda said. Partnering often enables the company to move faster.
Today, a big focus is “an agent-to-agent ecosystem,” Bolda said, connecting agents doing things like discovery and shopping. As an example of what that might look like, Schulze described the virtual agents Expedia has long employed to chat with users.
They are “pretty straightforward — you ask this question, you get this answer,” he said. “And the technology has evolved now so much to where it’s much more open and fluid.”
That can be combined with Expedia’s content and data, which he called the company’s biggest strength, “to really help our travelers make informed choices, to help our partners find solutions.”
Important to Expedia’s overall AI strategy, Bolda said, is ensuring it’s built into the platform.
“We’re not bolting it on,” she said. “It’s woven throughout the entire platform.”
Schulze said the company tries to be as nimble as possible with AI, and encourages a lot of experimentation. “The interesting thing is, as advanced as it is, I’d say we’re still in the early days,” he said.
The New York event was Expedia’s first local Explore event. For more than 20 years, Expedia has held an Explore event, typically in Las Vegas. Recently, the company decided to switch off years, alternating between Las Vegas and its headquarters in Seattle, with the addition of regional events. More regional events will be held in London, Cancun and Bangkok.
Gain hands-on experience with machine learning, AI, and big data tools in a career-focused online Master’s in Applied Data Science. Benefit from a top-ranked, affordable program with small class sizes, supportive faculty, and flexible schedules.
Now is the right time to take the next step in your data science journey and Bay Path University is the perfect place. With a master’s degree in data science, you’ll combine technical knowledge with an ability to translate data into actionable insights and drive decision making.
MS in Applied Data Science Degree Program
100% online and only 10 courses
Designed for working professionals to complete in as little as one year
Combines the technical, analytical, and communication skills needed for success
Provides hands-on application of Python, SAS, R, SQL and Quicksight
Includes a new Generative AI: Foundations and Application course
Has small class sizes and a dynamic community
Features extraordinary faculty working in the field
Enrolling now for the October 27th start
Why Bay Path University
1:1 mentoring from faculty and program staff
Access to tutoring, library and academic resources
Career services support for professional growth
Financial aid is available
No GMAT or GRE required
Accredited by the New England Commission of Higher Education (NECHE)
You must be logged in to post a comment Login