When people think about artificial intelligence, they often picture sleek start-ups or futuristic labs. But what happens when AI meets a company that has been innovating for over 100 years?
Unilever is one of the worlds largest consumer goods companies, home to brands like Dove, Hellmanns and Vaseline, with products used by 3.4 billion people every day. And behind those everyday items is a deep and evolving commitment to science.
From soap and margarine in the early 20th century to todays breakthroughs in sustainable packaging and personalized skincare, Research and Development (R&D) has always been our engine of progress. But now, that engine is being transformed by AI.
AI is not just a new tool in our labs, it is a new way of thinking. And for a company with a centurys worth of scientific data, that is a game-changer.
AI is reshaping every industry, but the companies that will be the most successful are the ones that know how to adapt, learn, and build on what they already know. While many legacy companies are exploring how to modernize through AI, the real opportunity lies in how they harness their institutional memory: the decades of research, product development, and consumer insights that can often sit untapped. This requires deep domain expertise, robust data stewardship, and a culture that values learning as much as legacy. When those elements align, AI can become a catalyst for transformation, by revealing the full potential of what has come before.
Unilever was born in the Victorian era, shaped by the industrial and scientific revolutions. Over the decades, we have evolved by responding to cultural shifts; from the transformation of domestic life in the mid-20th century to todays shifting expectations around skin health, beauty, and wellbeing to the growing urgency of sustainability. When new materials like Formica and stainless steel became common in mid-century kitchens, our scientists developed products tailored to these surfaces. This was not just chemistry, it was a scientific response to a changing way of life.
That same mindsetscience in service of real lifestill drives us today. But the questions were asking have become more complex: How do we support the skins natural microbiome? How do we clean homes without disrupting the ecosystems that live on our surfaces? How do we design products that are both effective and sustainable? These are not simple problems, and they require new ways of doing science. Thats where AI comes in.
With machine learning, we can uncover patterns that would take human researchers hundreds of years to detect. We are using AI to understand how microbes interact with our products, how skin responds to environmental stressors, and how we can personalize formulations for different needs and regions.
But here is what makes our approach uniquewe are not starting from scratch. Like many legacy companies our R&D archives stretch back over 100 years. We have records of every formulation, every trial, and every consumer insight. This historical depth gives our AI models something incredibly rare: context. While many companies are just beginning to build their data sets, established companies like ours are standing on a foundation that has been carefully constructed for generations.
Our scientists can unlock proprietary knowledge that was once siloed, scattered across teams, or locked in an archive. A century of skincare expertise is now structured, searchable, and ready to be applied. We are using AI to connect the dots across decades of research, accelerating discovery in new materials while simultaneously optimising formulations for specific needs, like different skin types. Were moving from research and discovery to formulation design and refinement in a single, integrated process, helping us respond faster and more precisely to peoples needs around the world.
This is not about replacing scientists with algorithms. It is about creating the conditions where human talent can thrive. Agentic AI systems give our teams the ability to ask better questions, explore more possibilities, and unlock insights from our data. By amplifying human creativity and empathy not automating it, were enabling our scientists to focus on what they do best: imagining, experimenting, and designing products that meet real human needs.
So why should this matter to anyone outside Unilever?
Because it shows what is possible when legacy meets learning. In an era where AI is reshaping every industry, the companies that thrive will not just be the newest or the loudest, they will be the ones that know how to adapt, how to learn, and how to build on what they already know.
AI rewards data maturity. It rewards curiosity. And it rewards companies that see technology not as a threat to tradition, but as a way to reimagine it.
We do not have all the answers. But we have learned that staying curious, being a learn-it-all not a know-it-all, is what keeps a company relevant for a century. AI is helping us stay curious at scale.
We believe the next 100 years of innovation will be driven by companies that embrace the partnership between human talent and agentic AI: hybrid systems that augment creativity, empathy, and scientific intuition. This is not just a story about technology. It is a story about legacy, learning, and the enduring power of science to shape the everydayonly now with a little help from artificial intelligence.
Alberto Prado is global head digital & partnerships at Unilever.
In a world where leadership is often mistakenly understood as a position of control, Ive found that true impact comes from serviceleading not from above, but alongside your team to achieve success.
This point was driven home about 15 years ago, when I made a resolution to read more books. Since starting this journey, one book stood out and shaped my leadership style more than any other: The Way of The Sheperd. It resonated with me so much that I even named my youngest son, Shep, after it.
This book takes readers through core principles of servant leadership and outlines a leadership plan that showcases how true influence comes not from authority, but from trust, empathy, and a genuine devotion to the growth and well-being of your team. I reread this book once a year to remind myself that the most effective leaders are ones who lead with compassion, understand what motivates each team member, and empower everyone to hold themselves accountable.
It underscores how a leader cant manage what they dont know, and how too many well-intentioned leaders focus solely on performance rather than their people. These lessons have influenced the way I lead, inspiring me to prioritize building trust and authentic connections with our team. At Kendra Scott, I make it a point to ensure everyone has direct access to memy email is always openand I encourage team members to reach out or schedule time to discuss anything on their minds.
Company connections
Weve also established a tradition of celebrating the meaningful connections within our company. For the past 9 years, weve passed down the shepherds staff. The tradition involves the current holder receiving the staff and keeping it for a set period of time before nominating someone else at the next family meeting. They share how this person has impacted them, add a meaningful token to the staff, and pass it on to the new nominee. The cycle continues from there. This simple yet powerful practice has reinforced our culture and the importance of the connections we share with one another.
However, this modern leadership philosophy wasnt one that always came naturally to me. Growing up, my life was all about sports, which helped me become the first in my family to go to college. The competitive mindset required for sports can sometimes be too focused on the individual. But the real lesson I took away, doing whatever it takes with your team to win, shaped my early approach to leadership.
3 things I know
As I moved into the business world and started working with founders like Ralph Lauren and Kendra Scott, I began to see how personal their connection to their companies was. That shifted my approach to leadership, and I started to focus more on building trust and creating a more supportive, nurturing environment for my team.
Here are three things I know now:
Fostering a sense of purpose leads to stronger performance
Investing in your team and infusing every position with importance helps to instill passion and purpose in your employees. Every employee should be empowered to feel as if they represent a brands vision. This authentic connection fuels productivity and drives success.
Making mistakes can be your biggest asset
You learn the most through your mistakes. In those moments, the best lessons come from acknowledgement and accountability. As a leader, being transparent and vulnerable about your own missteps can set a tone that not only enhance company culture, but also make employees feel secure in taking risks.
The importance of leading with a learners heart
Great leaders are great learners. No one has all of the answers, regardless of their position in an organization. Thats why its important to maintain a culture of continuous learning and collaboration. The more you learn, the more tools and opportunities you have.
In the end, leaders cant be successful without their team, and leadership isnt about control, power, or having all the answers. Its about showing up for your people, creating space for them to grow, and demonstrating humility through it all. Whether its by reading a book about servant leadership or establishing supportive methods, Ive learned that true leaders are those who cultivate intentional connections with their employees. There is still much to learn, and my journey is still unfolding, but one thing I know to be true is that leading with trust and empathy is the type of legacy worth leaving.
Tom Nolan is CEO of Kendra Scott.
Amaras Law, coined by the American scientist and futurist Roy Amara, says humans tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run. If the first half of 2025 is anything to go by, in the AI era, the runs are getting shorter, and the effects of the technology will be larger than weve seen in a generation.
In a matter of months, the conversation in companies has accelerated far beyond if AI is a useful productivity tool, to where and when it can be applied. Across industries and geographies, executives are acknowledging that AI is a general-purpose business solution, not just a technical one.
Despite widespread workplace adoption, the focus on cybersecurity has not kept pace. In the rush to adopt AI systems, applications and agents, companies are failing to consider that rapid deployment of these new technologies could lead to data breaches and other security risks.
That matters because AI models are not only getting more powerful but also more useful for enterprises.
More enterprises are using AI agents
As of early June, OpenAIs base of paying business users reached 3 million, up from 2 million in February. In a move for that market, ChatGPT can now connect to popular business apps such as Google Drive, Dropbox, and Sharepoint, allowing workers to quickly access answers that are locked in dispersed documents and spreadsheets.
Confusion, and even fear, about AI agents has given way to exploration and adoption. Among US-based organizations with annual revenues of $1 billion or more, 65% were piloting AI agents in the first quarter of this year, up from 37% in the space of a single quarter.
Microsofts Azure AI Foundry, its platform for building AI agents, processed 100 trillion tokens in the first three months of 2025 (with one token representing the smallest unit of text that an AI model processes)a five-fold increase year-on-year. At the same time, the cost per token more than halved, spurring higher use and creating virtuous cycles of innovation.
As John Chambers, the former CEO of Cisco, says, AI is this generations internet revolution but at five times the speed, with three times the outcome. Beyond the hype that haunts the sector, there are signs of enterprise AI adoption everywhere.
In his latest letter to shareholders Alex Karp, CEO of Palantir Technologies, describes a ravenous whirlwind of adoption of AI. IBM, which has rolled out its AI strategy to 270,000 employees, reports that AI already handles 94% of routine human resources tasks.
At Shopify, the e-commerce group, AI usage is now a baseline expectation, CEO Tobias Lütke said in an employee memo. The workplace automation company Zapier, which took steps to embed AI across its workforce, says that 89% of employees actively use AI in their daily work.
The list goes onand its not just technology companies. JP Morgan, the worlds largest bank, has rolled out GenAI tools to 200,000 staff members, and says employees have each gained one-to-two hours of productivity each week.
AI acquisitions are plentiful
The shift from novel to mass-market tech is reflected in the business strategies of the main AI model makers, which are reimagining themselves as application companies. In the space of two weeks, OpenAI, the ChatGPT parent, appointed a CEO of Applications and then acquired IO, the AI device startup founded by former Apple designer Jony Ive, for $6.5 billion.
Meta, perceived to be behind in the AI race, has invested $14.3 billion in Scale AI, which provides data and evaluation services to develop applications for AI. Meanwhile, Apple is reported to have had internal talks about buying Perplexity AI, a two-and-a-half year-old AI model maker.
AI app security is rarely discussed
Companies are naturally focused on the potential and performance of AI systems, but it’s striking how rarely security is part of the story. The reality is that the speed of deployment of AI apps and agents is leaving companies at risk for breaches, data loss, and brand impact.
For example, an AI system or agent that has access to employee HR data or a banks internal systems leaves a company open to possible cyberattacks by bad actors. In business-critical applications, risks emerge at every stage of the development cycle, from choosing which AI model to use and what systems to give it access to, right through to deployment and daily use.
In our work on testing the security of AI models with simulated attacksknown as red-teamingand creating the CalypsoAI Model Security Leaderboards, we have discovered that, despite performance improvements, new or updated AI models are often less secure than existing ones. At the same time, existing models can see their security score slip over time.
Why? Because the attacks keep progressing and bad actors learn new tricks. More techniques and capabilities of breaking or bypassing AI model securities keep being invented. Simply, the attack techniques are getting better and they’re causing AI models that have only recently launched to become less secure.
That means that organizations that begin using an AI system or agent today, but don’t stay up to date with the latest threat intel, will be more vulnerable as attack techniques increase in capability and frequency. As corporate AI systems gain autonomy and access to sensitive data, what is safe today may not be safe tomorrow.
The research firm Gartner has forecast that 15% of day-to-day business decisions will be made autonomously by agents by 2028, though that percentage may increase by then. Against that backdrop, virtually all the security protocols and permissions in enterprises are built for human workers, not for AI agents that can roam through company networks and learn on the job.
That mismatch opens up vulnerabilities, such as the ossibility of agents accessing sensitive information and sharing it inappropriately. Poorly secured agents will be prime targets for hackers, particularly where they have access to valuable data or functions such as money transfers. The consequences include financial loss and reputational damage.
Final thoughts
Securing these new systems will be critical to AI adoption and to successful return on investment for the companies involved. A new security paradigm, using the capabilities of agentic AI to secure enterprise AI, is needed to allow innovation to thrive and agents to reach their potential.
While the development of AI models and systems so far can reasonably be summarized as better, cheaper, less secure, the final part of that equation must improve significantly as the emerging application-first AI era accelerates. Once that happens, Roy Amara seems certain to be proven right once again.
Donnchadh Casey is CEO of CalypsoAI.
As Zohran Mamdani declared victory in the New York City Democratic mayoral primary on Tuesday night, one had to wonder: Has anyone checked on the finance bros?
On X, the Wall Street meltdown was already well underway. It appears that NYC is electing to commit suicide by Mayor, wrote Jim Bianco, president of Bianco Research. Its officially hot commie summer, added Dan Loeb, CEO of hedge fund Third Point and longtime Cuomo backer.
every NYC finance guy on my main feed right now pic.twitter.com/Z7anouAFio— (@EffMktHype) June 25, 2025
Loeb wasnt alone. Billionaires like Michael Bloomberg and Bill Ackman had backed Andrew Cuomo, still seen as the frontrunner even in the races final days. Bill Ackman drafting his thoughts on Mamdani rn, one post joked, alongside an image of an essay-length text being written. Another great part of Mamdanis victory is that it means Michael Bloomberg pretty much lit over $8 million on fire for no reason lol, another X user wrote.
Bill Ackman drafting his thoughts on Mamdani rn pic.twitter.com/e7pCMMKKO4— litquidity (@litcapital) June 25, 2025
The finance industrys reaction isnt surprising. A Mamdani win in Novembers general election could bring what Wall Street dreads most: tax hikes and tighter regulations threatening corporate and investment interestsfueling the familiar cry of a wealthy exodus. Wealthy New Yorkers moving to Miami if Zohran wins, one Instagram meme joked.
View this post on Instagram A post shared by New Yorkers (@newyorkers)
Lets not forget the finance bro who did vote for Mamdani (starter pack includes a Carhartt beanie and a copy of The Communist Manifesto). NYC girls with trust funds were calling him Zaddy Zohran and you thought he was going to lose? one user posted on X.
pic.twitter.com/0mZY1Kfw2s— Overheard on Wall Street (@OHWallStreet) June 24, 2025
One post, acknowledging defeat, featured an AI-generated image of fleece-clad finance bros scanning groceries at a city-run market: Well boys, onto the new bullpen.
Well boys, onto the new bullpen. pic.twitter.com/GUxzYaYZw6— Overheard on Wall Street (@OHWallStreet) June 25, 2025
Another post on X perfectly summed up the general mood: Investment bankers reacting to NYC nominating a socialist for Mayor. The accompanying caption? wed all be fine with a lot less, right?
Investment bankers reacting to NYC nominating a socialist for Mayor pic.twitter.com/4SY66q4qTn— litquidity (@litcapital) June 25, 2025
In my conversations with business leaders around the world, I consistently hear the same phrase to describe what they want to achieve for their workforce: AI fluency.
I often tell them that to achieve AI fluency, we need to treat it as a foreign language.
Like learning a new language, becoming AI fluent requires dedication, immersion, and practice. Fluency transforms how we think and communicate. Becoming fluent requires us to overcome the fear of making mistakes or incurring risks.
Yet theres one crucial difference between achieving fluency in AI versus a new language: When learning a new language, we step into an established culture. With AI, were learning the culture while simultaneously creating it.
The big question: How can organizations build these cultures and become laboratories of AI fluency? Here are three ways to foster AI fluency.
1. Create an immersive environment
Whether were learning Spanish, Mandarin, or any of the other 7,000 languages in the world, immersion is an essential step to fluency. Living where the language is spoken forces you to adapt, to think differently, and to develop new neural pathways.
AI requires the same commitment. Organizations are uniquely positioned to create these immersive environments where employees interact with AI tools daily, not as occasional novelties but as essential components of their workflow. From Udemys work with thousands of organizations around the world, helping to create these environments, weve found that organizations succeed when they integrate AI across departments, from marketing teams using generative AI for content creation to HR departments employing AI-powered skills assessments.
Immersive environments are built when employees understand they need to become fluent to reach their goals. That means the most successful AI adoption happens when tools directly address employees pain points. Just as language learners progress faster when they need the right words to order food or navigate transportation, employees embrace AI more readily when it solves real problems they face.
Organizations seeking AI fluency must balance structure with exploration. Consider how language learning works: Structured lessons provide grammar and vocabulary, but real learning happens through conversation and experimentation.
Similarly, building organizational AI fluency requires a few basic building blocks:
Upskilling on foundational AI capabilities and limitations, like learning the rules of grammar.
Creating a sandbox-style environment where people can experiment without fear of consequences.
Developing communities of practice where people can find social support to troubleshoot, ask questions, celebrate successes, and motivate each other to keep experimenting.
Establishing guidelines for when to rely on human judgment versus AI, how to evaluate AI outputs, and how to maintain human connection in AI-mediated interactions.
2. Overcome fluency barriers
The barriers to AI fluency mirror those of language learning. Fear of embarrassment prevents many language learners from practicing conversation, just as fear of looking incompetent may prevent employees from experimenting with AI. Imposter syndromethe feeling that everyone else knows more than you doimpacts both AI and language fluency.
The solution is creating psychologically safe environments where questions are welcomed, and mistakes are treated as learning opportunities. Leaders like Salesforce CEO Marc Benioff model this by encouraging employees to approach new challenges with a beginners mind, getting curious instead of expecting immediate mastery and understanding.
Whats more, both language learners and AI adopters often experience an uncanny valley stage where they know enough to recognize their limitations but not enough to feel confident. Supporting people through this phase is critical. This is where many abandon the journey if theyre not properly encouraged. In this case, encouragement can come not only from leaders, but from the environments leaders create such as building supportive communities of practice where learners can share their struggles with gaining fluency. This normalizes the experience, while reminding them that this uncomfortable stage is not just common but also a sign of meaningful growth.
3. Create culture while learning
This is where the language metaphor ends.
While becoming AI fluent, were simultaneously students and architects of the culture. This dual role presents unprecedented responsibility and opportunity.
Leaders must consciously shape how AI integrates into the organizational culture by establishing rules and norms that preserve human creativity and connection while leveraging AIs capabilities. This means modeling thoughtful AI usage, celebrating innovative applications, and continuously reinforcing that AI serves human objectives, not the reverse.
The organizations that thrive will be those that build immersive environments where employees can become AI fluent and build cultures where technology amplifies uniquely human capabilities. In a workplace where managers offload administrative or basic creative tasks to AI agents, employees would gain hours back in their day. This would allow them to spend more time coaching their teams, helping them solve problems, identify opportunities for growth, and learn the best ways to motivate them during times of change and upheaval.
The journey to this future begins with recognizing that AI, like any language, isnt just a skill to acquire but a new way of thinking.
Hugo Sarrazin is the CEO of Udemy.
Donald Trump is no stranger to selling unusual merch. The president has been known to slap his likeness on Bibles, guitars, and gold-plated phones if itll bring in a bit of cash. Just last year, the Trump Organizations licensing deals netted Trump tens of millions of dollars in profits. But today, Trump has outdone himself with what has to be his strangest piece of merch ever: a T-shirt featuring his own mug shot captioned with the word Daddy (in call caps).
The official Trump Daddy shirt, featured in an email newsletter sent to subscribers on June 26 from the Trump National Committee JFC, appears to be a limited-time offering separate from the official Trump store. Its being sold through its own webpage with the notice These are going to sell FAST. HURRY, claim your shirt today, as well as the somewhat more unsettling all-caps call to action CLAIM DADDY SHIRT.
[Screenshot: Winred]
The merch-ification of Trumps mug shot is a fairly tired tack at this point, given that the president has already sold pieces of the suit worn in the image, reprinted it on rally posters, and used it as inspiration for his official White House portrait. Earlier this month, Cara Finnegan, author of the book Photographic Presidents: Making History From Daguerreotype to Digital, told Fast Company that the mug shot has been used so often by Trumps administration that it arguably is his presidential portrait.
However, the use of the moniker Daddy is altogether new. It seems to be a nod to a comment made June 25 by Mark Rutte, secretary-general of NATO, who appeared to give Trump the title when he said, “Daddy has to sometimes use strong language. The media (and, now, it seems, Trumps team) took the comment as a reference to a viral moment earlier in the week when Trump casually dropped the f-bomb on camera.
Rutte has since denied that the term was directed at Trumpbut, unfortunately, that didnt stop the White House from posting an edit of the President set to Ushers sexual innuendo-laden R&B song, “Hey Daddy (Daddys Home). Trump himself appeared to endorse Ritters use to the term daddy at a later press conference. Now, it seems, the Trump National Committee is shortening the life cycle of real news-turned meme-turned merch. Mere hours after the initial stir around the daddy comment, its already a shirt.
Its not entirely shocking to see Trumps fundraising committee capitalize on this moment, given that his administration has already thrown convention to the wind by constantly glamorizing his mug shot. Still, this development is markedly more unusualand if this flies as a kind of Presidential merch, we dont want to know what might appear on a Trump shirt next.
A first public test of robotaxis by Tesla in Austin, Texas led to multiple traffic problems and driving issues, videos from company-selected riders showed over the first few days.
Chief Executive Elon Musk has tied Tesla’s financial future to self-driving technology, and with Tesla sales down, the stakes are high. He said Tesla would roll out the service to other U.S. cities later this year and predicted “millions of Teslas” operating “fully autonomously” by the second half of next year.
The Tesla fans invited to the trial were strongly supportive and posted videos of hours of trouble-free driving, but issues drew questions from federal road safety regulators and auto safety experts.
Issues included Tesla robotaxis entering the wrong lane, dropping passengers off in the middle of multiple-lane roads or at intersections, sudden braking, speeding and driving over a curb.
In one instance, a robotaxi drove into a lane meant for oncoming traffic for about 6 seconds. It had pulled into an intersection in its left-turn lane with its turn blinker on. Then the steering wheel wobbled momentarily, and instead of turning it proceeded straight into the lane meant for oncoming traffic, prompting a honk from a car behind it.
In another incident, the car suddenly braked with no obstruction apparent in the video. The passenger jerked forward and their belongings were thrown to the floor. In a third video, taken from another vehicle, a robotaxi abruptly stopped twice in the middle of the road while passing police vehicles with flashing lights.
Tesla is conducting the test with human safety monitors in the front passenger seat. A fourth video showed the safety monitor hitting a button to stop the robotaxi when a delivery truck in front of it started backing up.
“This is awfully early to have a bunch of videos of erratic and poor driving,” said Philip Koopman, a Carnegie Mellon University computer-engineering professor and autonomous-technology expert. “I was not expecting as many videos of problematic driving on the very first day,” he said. Tesla is testing about 10 to 20 robotaxis, which are standard Model Ys with advanced software, and has been giving rides since Sunday afternoon.
Reuters was able to independently verify the locations of at least 11 videos showing issues. Tesla did not respond to a request for comment.
A City of Austin spokesperson said officials are aware of the Tesla issues documented on social media and that “when a potential legal or safety concern is brought to our attention, we promptly share it with the company.” The spokesperson added that the police department is “actively collaborating with Tesla” to ensure officers can safely interact with the robotaxis.
‘Caught on camera’
The incidents caught on camera did not involve accidents, and one expert said some reflected a decision to focus on safety.
“So far so good. It handled the situations very well and likely better than even good drivers,” Alain Kornhauser, Princeton University professor of operations research and financial engineering, said by email. He added that it would be more dangerous to drive at less than the speed of prevailing traffic, for instance.
Tesla’s experiment is unusually public. Other companies faced similar issues: Alphabet’s Waymo and General Motors’ Cruise had their own share of traffic mishaps after showing up on Austin streets. City officials logged dozens of instances over the past two years where residents and authorities reported that robotaxis blocked traffic by stopping in the middle of roads, failed to respond to police directions and could not deal with emergency vehicles and road closures.
A serious accident involving a pedestrian in 2023 led Cruise to shut down last year. Waymo is the only robotaxi service in the U.S. to ferry paying customers without a human backup driver or in-car safety monitor. It started offering rides to the general public through Uber in Austin earlier this year.
Musk for years has failed to deliver on promises that self-driving Teslas are just around the corner. Tesla rolled out the service for a flat fee of $4.20 to a limited number of handpicked riders. The service is not available to the broader public and the robotaxis operate in a limited area, and avoid difficult intersections and bad weather.
Riders were rarely bothered much by driving issues. Farzad Mesbahi, a former Tesla program manager, and his co-passenger hit the “drop off early” option during a ride. The vehicle stopped in an intersection with a stoplight, his video showed. They exit quickly and walk to the sidewalk. “The car should have known to not stop there,” Mesbahi is heard saying after the ride. “Opportunities for improvement,” the co-passenger says.
That is an example “most companies would not be comfortable with,” said Kara Kockelman, a professor of transportation engineering at the University of Texas at Austin, adding that she was surprised by the traffic mistakes.
“Dropping off people in the middle of a six-lane road or edge of a busy intersection when the traffic is going in the opposite direction is pretty dangerous. They definitely did not want to do this or be caught on camera,” she said.
Abhirup Roy, Rachael Levy, and Chris Kirkham, Reuters
Additional reporting by Inaki Malvido, Fernando Robles, and Richa Singh.
When it comes to AI agents, the makers of QuickBooks are hoping that youre into it.
Intuitthe fintech platform that owns TurboTax, Credit Karma, Mailchimp, and QuickBooksannounced that it has implemented a new set of AI agents into its products. The company showcased how the AI agents work within QuickBooks at an event on June 24, and Fast Company was able to see a demonstration of how the agents can help business owners and entrepreneurs use them to speed up their bookkeeping and accounting processes.
QuickBooks will incorporate a Payments Agent, an Accounting Agent, a Customer Agent, and a Finance Agent, all of which are designed to become intimately familiar with a businesss specific customer base and financial track record, offer up insights, and make additional analyses. And though its just now being rolled out, the new AI capabilities have been in the works for a long time.
This is five or six years in the making, Sasan Goodarzi, Intuits CEO, tells Fast Company. Weve made huge investments in the past five years, he says, and the company has taken its time because when it comes to bookkeeping and accounting, accuracy matters.
[Image: Intuit]
In other words, Goodarzi says that while an AI tool like ChatGPT might spit out wrong or incorrect information, a customer relying on QuickBooks to crunch their numbers needs to be absolutely sure and trust in Intuits accuracyotherwise, they could find themselves with serious issues. As such, Intuit wanted to make sure everything was above board before launching to its full customer base. If it screws up, its a big problem, he says.
Additionally, Goodarzi says that business owners are relying on a huge number of apps and platforms to run their companies, an issue that Intuit is trying to simplify. What Im hearing from customers all the time is that theyre over-digitized, there are too many apps. Theyre not getting the benefit from their time and money, he says. This is about creating a one-stop shop, a refreshed way to discover all of the capabilities within QuickBooks, he says, noting that many of Intuits customers are unaware of how many tools exist within the QuickBooks ecosystem.
And its the discovery and engagement with those tools that Goodarzi says has been the area of the most positive feedback.
But the primary question: Are the AI implementations actually producing value for users? Yes, Goodarzi says. He notes that during the testing phase, the new AI capabilities have led to significant time and money savings for users, though that can be difficult to quantify, and expects that the new features will both resonate with QuickBooks wider user base when they officially launch on July 1, and help the companys bottom line.
I was talking about AI changing the world six years ago, and people were laughing at me, he says. Now, were actually seeing natural adoption, driving incredible value.
Anna Wintour, an icon in the fashion industry who has been at the helm of Vogue for nearly four decades, will be stepping down as editor-in-chief, according to a number of publications including WWD and Business of Fashion.
Wintour told staffers on Thursday the company is seeking a new head of editorial content at American Vogue, per USA Today. She will remain as Condé Nast’s chief content officer and as Vogue’s global editorial director.
She is best known for her impeccable sense of fashionimmortalized by her famous bob hairstyle, large sunglasses, and Georgian Collet necklacesas well as for helming the Met Gala, considered one of fashion’s most exclusive social events.
As a businesswoman, she has been criticized and praised for her direct, take-no-prisoners “girl boss” management style, which was immortalized in the 2006 movie The Devil Wears Prada, in which her character was played by Meryl Streep.
Wintourwho is British, but become a U.S. citizen in 1972 after marrying American child psychologist David Shafferhas presided over Vogue since 1988, making it the “fashion bible” that it is today.
As chief content officer, Wintour oversees a number of Condé Nast brands, including: Wired, Vanity Fair, GQ, AD, Condé Nast Traveler, Glamour, Bon Appetit, Tatler, World of Interiors, Allure, and others, except The New Yorker, which editor David Remnick runs.
Want more housing market stories from Lance Lamberts ResiClub in your inbox? Subscribe to the ResiClub newsletter.
Since the Pandemic Housing Boom fizzled out, major homebuilders across various marketsespecially in top pandemic boomtownshave had to cut net effective home prices to avoid a deeper sales pullback. However, some builders, like Lennar and D.R. Horton, have primarily done so through larger incentivessuch as mortgage rate buydownsin part to protect community comps and avoid upsetting buyers already in their backlog.
Speaking to analysts on Tuesday, KB Homewhich prefers outright home price cuts over incentivessaid that some buyers turning to some of their competitors are effectively overpaying for new builds just to get rate buydowns, and if they need to sell in the immediate future, they might not be able to fetch the artificially high base price they paid.
I believe that there are customers [of other homebuilders] that are overpaying for the home to effectively get an incentive. So they’re tied into this higher price that they’re gonna be stuck with forever until they sell that home. They may potentially be upside down when they try to sell that home versus a clean, simple, transparent way of sellingthe value of what we offer, KB Home COO Rob McGibney said on the builders June 23, 2025 earnings call.
Below are ResiClubs other takeaways from KB Homes Q2 earnings report and earnings call this week.
KB Home says the 2025 housing market is softer than expected
KB Homes net new orders by Q2:
Q2 2018 > 3,532
Q2 2019 > 4,064
Q2 2020 > 1,758 (COVID-19 lockdowns)
Q2 2021 > 4,300
Q2 2022 > 3,914
Q2 2023 > 3,936
Q2 2024 > 3,997
Q2 2025 > 3,460
The actions we began to take late in our 2025 first quarter, evaluating base pricing in every community relative to local market conditions, then repositioning our communities with a focus on offering the most compelling value, led to strong net orders in March. However, our net orders declined in April and May, which did not follow the typical spring trajectory, said KB Home COO Rob McGibney.
McGibney added that: As a result, even though our average community count was in line with our projection, and our cancellation rate was fairly steady, our monthly absorption pace per community was 4.5 net orders compared to 5.5 in last year’s second quarter. While our net order pace was below our internal goal, we believe it ranks high among the large production homebuilders.
KB Home: All of the markets we operate in experienced some level of softening at some point during the quarter
While the pricing story continues to be very local and vary a great deal across the country, most markets are at least seeing some softening.
I would say that all of the markets we operate in experienced some level of softening at some point during the quarter, KB Home COO Rob McGibney told investors on Tuesday. Markets that I would say where we’re still seeing relatively strong demand and sales performance would be Las Vegas, the Inland Empire, the North Bay in Northern California, Texas markets like Houston and San Antonio.
McGibney added that: By contrast, some of the markets that are facing some more significant headwinds recently are like Sacramento and Seattle. They’ve slowed down a little bit, and we’ve had to do a little more there with price relative to some of the others. Markets like Austin and Colorado, Jacksonville, Orlando, and Florida [have been weaker too]. Places where resale supply has increased and starts putting pressure on pricing and creating more competition and just more choices for buyers. But, you know, it is very local, very specific, [we] can’t put a market condition on an entire state or even an entire market in most cases, it’s community by community.
KB Home had to make some bigger price cuts in markets where resale inventory is above 6 months
On Tuesday, KB Home told analysts that it cut base home prices in half of its communities in the quarter ending May 31st.
In the markets where you’ve seen resale inventory or resale supply get back to norms or above those norms of six or seven months of supplythose resales become a more formidable competitor than they were to us back when we would measure months of supply in terms of weeks instead of months. And on the flip side, most of the markets where resale supply has stayed fairly suppresed and limited, we’re tending to see better results there, KB Home COO Rob McGibney told analysts on Tuesday.
Margin compression continues
During the Pandemic Housing Boom, many publicly traded homebuilders achieved record profit margins as home prices soared and buyer demand ran red hot. Ever since the national housing demand boom fizzled out in the summer of 2022, many large homebuilders have reduced margin and made affordability/pricing adjustments where and when needed to maintain their sales pace or prevent a bigger sales pullback.
That includes KB Home, which reported a housing gross profit margin of 19.3% in Q2 2025or 19.7% excluding inventory chargesdown from a cycle peak of 26.7% in Q3 2022. Its margin has now compressed all the way back to pre-pandemic 2019 levels.
KB Home: Only two minor price increases [related to tariffs] to date
So far, tariffs havent had much impact on KB Homes material costs.
Homes that we started in May came in at the lowest cost per square foot year to date, as our divisions are continuing to drive better performance on cost. Our costs, including lumber, are protected for almost all of our third-quarter starts under the terms of our supply contracts. Our national purchasing team, working with our divisions, has done an excellent job holding off tariff-related cost increases, with only two minor price increases to date, KB Home COO Rob McGibney told analysts on Tuesday.
