Editor’s note: This is AI Impact, Newsweek’s weekly newsletter where each week, we will explore how business leaders are unlocking real value through artificial intelligence.
Tap or click here to get this newsletter delivered to your inbox.
Signal Capture
Signals from the front lines of AI adoption
Is Your Company Built for AI?
By Adam Mills
At a time when much of the AI conversation is still centered on experimentation, Amit Kumar is seeing early signs that companies are beginning to rethink how work actually gets done.
“We’re starting to see a movement from POCs to actually meaningful process changes,” Amit Kumar, managing partner and global head of consulting at Wipro, a technology services and consulting company, told Newsweek.
For much of the past two years, companies have focused on pilots, proofs of concept, and isolated deployments. That phase is now giving way to a more difficult reality: putting AI into day-to-day operations in a way that holds up under scale.
The limits are increasingly organizational.
“The bottleneck is not AI capability,” Kumar said. “It is organization capability.”
The gap between what AI can do and what companies are prepared to absorb has become a defining challenge. While tools have advanced quickly, many organizations still rely on fixed processes, legacy governance, and roles that were never designed for intelligent systems. The result is a widening gap between visible activity and measurable outcomes.
“There’s a recognition that this is real, that this is not something five years out,” Kumar said. “What they are struggling with is recognizing that to get value out of this, you also need to change.”
That reality is increasingly shaping how companies frame the problem. A year ago, Kumar said, clients were asking where AI could be applied and which use cases to test. Now, the question is more direct: “Which part of the business can we run with AI?”
That mindset is also reshaping expectations for consulting firms.
“It’s not an evolution,” Kumar said. “It’s a reinvention of the consulting industry.”
Traditional models built around scaling headcount are giving way to approaches that embed AI directly into delivery. Wipro has also used itself as a proving ground, reworking internal workflows and introducing AI agents across functions such as financial reconciliation, recruiting and deal structuring, running them alongside human teams to compare output and accuracy over time.
The aim is not to remove human judgment, but to rethink how it is applied.
“The biggest barrier is change readiness,” Kumar said. “Are you ready to adopt this change?”
That question is increasingly less about infrastructure and more about behavior. While many organizations have made progress on data and cloud foundations, adoption often stalls at the individual level. Employees remain uncertain about how AI will affect their roles, while leaders work through how to redesign workflows, incentives and accountability.
“The anxiety is not about change,” Kumar said. “It is about the speed of change.”
Overcoming that friction requires a different approach. Rather than treating AI as a top-down initiative, Kumar said companies need to embed it in everyday work and let employees experience its value firsthand.
“Make it personal,” he said.
That bottom-up familiarity, combined with changes to governance, roles and decision-making, is what enables AI to move beyond isolated deployments. The companies that succeed in doing that, Kumar said, will quickly pull ahead.
“The next phase of competition will not be who has AI,” he said. “It will be who can run their business with it.”
Core Intelligence
Former Genpact CEO Says Software Companies Must Own the Outcome
By Adam Mills
The old software bargain depended on a clean handoff: vendors sold the product, and someone else made it work inside the customer’s business.
AI makes that arrangement harder to defend because the product is no longer just helping someone complete a discrete function. Increasingly, it is being pulled deeper into the customer’s operating process.
“No longer am I going to buy tasks, I want to buy work,” Tiger Tyagarajan, former CEO of Genpact and now a senior advisor and board member across technology, services, private equity and venture capital, told Dr. Ranjit Tinaikar, the former CEO of Ness Digital Engineering and the host of Newsweek’s “AI Impact Forum” series, during the webinar.
If customers are buying work rather than tasks, they will expect vendors to stand closer to the result. A product that helps a sales, marketing or customer service team complete a function may no longer be enough if the customer’s real expectation is a business outcome.
“For the first time, software companies have to start promising outcomes,” Tyagarajan said. “AI is probabilistic. It’s not deterministic.”
AI systems do not behave like traditional software, in which the same input reliably produces the same output. They require monitoring, governance and adjustment after deployment, making services harder to treat as a separate layer added later.
Software companies have often protected their margins by keeping that work at arm’s length. Partners could implement, customize and support the product while the vendor focused on selling high-margin software. AI weakens that separation because the quality of the product increasingly depends on how well it performs inside live workflows.
“The moment that changes to rapid, you need to bring it closer,” Tyagarajan said.
Services, in this context, become less like an add-on and more like part of the product experience. If an AI-enabled system fails to deliver the intended result, the customer may not care whether the problem lies in the software, implementation, or workflow design. The vendor’s credibility is tied to the outcome.
The issue is less whether software companies become services firms than whether they can continue to treat delivery as someone else’s problem. AI makes that harder because product performance, customer workflow and business results are increasingly bound together.
The companies best positioned may not be the ones with the broadest platforms or the largest services teams. They may be the ones that can learn fastest from customers, close the gap between product and delivery, and prove that their systems keep working after the contract is signed.
“No one says segregate it, because you can’t,” Tyagarajan said. “The world we are entering now is very difficult to segregate and you shouldn’t segregate.”
You can watch the full discussion and read the recap here: Former Genpact CEO: Don’t Be Afraid to Cannibalize Revenue With AI.
Upcoming Webinars
The Trillion-Dollar Question: Who Wins and Who Loses in the AI Services Economy?
As AI accelerates across the enterprise, it is forcing leaders to rethink how technology spend is structured—and how much productivity it must deliver. In an upcoming AI Impact Forum session, Dr. Ranjit Tinaikar speaks with Noshir Kaka, senior partner at McKinsey & Company, about how this is reshaping enterprise spending—and what it means for organizations trying to keep up.
Join the live discussion on Thursday, May 28, at 10 a.m. Eastern. Register, for free, right now.
Prompt Injection
What’s one recent insight you’ve learned about AI?
“It’s the exponential nature of the change that AI creates. That exponential change in capabilities means that functionality is coming at us at an accelerated rate – Mythos being the latest example. We are accustomed to linear change. But with AI we get surprised by the category shifts when it has matured enough to open whole new purposes, ideas, concepts and ways of doing things. These are going to come at us at increased speed. That’s the realization. Understanding that the limitations going forward are physical with numbers of data centers rather than technical means knowing that we are in for an interesting ride.
AI will change the way we structure our way of interacting with information, our work and our lives. Knowing that these changes are going to be coming at us at an accelerated rate means that we will need to be flexible about our understanding of the world if we are to stay current in a life with a relationship with technology.
Being involved in managing and financing risk that arises out of AI technology, it means that we must have the imagination to not only predict upcoming uses and purposes, but also the downstream risks from corporate and societal transformations. It opens whole new worlds and requires that we keep our minds open and welcome the surprises as they come.”
Have your own lesson to share? Email us at: a.mills@newsweek.com
Run Log
AI use case of the week
By Adam Mills
Streaming audiences are paying more to see fewer ads. Many viewers want uninterrupted content, but avoiding commercials often means paying more for another subscription tier.
The problem is also timing. Traditional product placement is usually arranged months before a show is finished, while brand campaigns are often planned just weeks before launching. By the time advertisers are ready, the opportunity to include them in the content has usually passed.
Omar Tawakol, chief executive officer of Rembrand, a software company that uses AI to insert brands into videos after production, is focused on that timing problem.
The system can add a brand as late as the day before a show airs, rather than during filming. Programming that once required early coordination can be updated just before release. A scene that would have aired unchanged can include a brand in a way that appears native to the original footage.
Technical flexibility created a standards problem. “Just because you can do anything doesn’t mean you should,” Tawakol said. As the approach expanded, attention turned to defining rules that would let content owners integrate brands without negotiating each insertion individually.
Rembrand’s bet is that finished video can still be updated, giving content owners another way to support programming without adding more interruptions for viewers.
Have an AI use case to share with us? Email us at: a.mills@newsweek.com
Context Window
■ Pirelli has acquired a 30 percent stake in Univrses, a Stockholm-based computer vision and AI company whose technology uses vehicle and tire sensor data to map road conditions and help improve driver-assistance systems, autonomous vehicles and infrastructure maintenance. [Newsweek]
■ A security review found more than 5,000 AI-built web apps with little or no authentication, exposing a new shadow IT risk as employees use AI coding tools to create software outside traditional development and security reviews. [WIRED]
■ Caltrans has launched a five-year pilot of AI-powered adaptive traffic signals along California’s Highway 68 corridor to test whether real-time traffic control can reduce congestion, improve public infrastructure planning and provide a lower-cost alternative to roundabouts. [SFGATE]
■ Tech companies are cutting middle-management roles and pushing remaining managers to act as hands-on “player-coaches,” with AI agents helping smaller teams take on more work and changing what leadership looks like inside flatter organizations. [Business Insider]
■ Anthropic has signed an agreement with SpaceX to use all compute capacity at its Colossus 1 data center, securing more than 300 megawatts of capacity and over 220,000 Nvidia GPUs to raise Claude Code and API usage limits and improve capacity for Pro and Max subscribers. [Anthropic]
■ A new study found that AI models could copy themselves across networked computers in a controlled test environment, though cybersecurity experts said the result is more a future warning sign than a near-term enterprise threat. [The Guardian]
Transfer Protocol
Tracking executive moves across the AI landscape
Ben Durham-Kilcullen, previously chief innovation officer at Stova, has been appointed Ondaro’s first chief AI officer, where he will lead the company’s artificial intelligence strategy and guide AI implementation, scaling and client delivery across the business.
Tim Finley, previously a go-to-market and strategy leader for data and AI services at Amazon Web Services, has been appointed senior vice president of AI at Couchbase, with responsibility for AI product strategy, roadmap development and the company’s internal AI transformation.
Viksit Gaur, previously an applied AI leader at Dropbox and founder of Myra Labs, has been appointed chief AI officer at Aura, overseeing expansion of the company’s intelligence platform for agent-led security and well-being monitoring.
Alex Katouzian, previously executive vice president and group general manager of mobile, compute and extended reality at Qualcomm Technologies, has joined Intel as executive vice president and general manager of client computing and physical AI, a role focused on aligning the company’s client computing business with emerging AI systems across robotics, autonomous machines and other devices.
Michel Combes, currently chairman and chief executive officer of Brightspeed and a former chief executive officer of Sprint, SoftBank International and Alcatel-Lucent, has been appointed chief executive officer at Lambda, where he will help scale the company’s AI infrastructure business through capital formation, AI factory expansion and operational growth.
Know someone on the move in AI? Send job change info to a.mills@newsweek.com
Magic Moment
What’s the most fun or unexpected way you’ve used AI lately?
“Running a lean team means we can’t afford to not use AI. At the same time, as lawyers, we can’t afford to use it carelessly, so I’ve become obsessed with building guardrails into the tools we use on a daily basis.
My current AI passion project has been creating an agent that automatically surfaces ethical traps as I review legal documents. I’ve been training it with guidelines issued by organizations like the American Bar Association and State Bar of Texas so it proactively flags issues – whether that’s citation hallucinations or potential breaches of client confidentiality. This added set of ‘eyes’ is invaluable. It’s like having a sharp, fourth-year associate who is dedicated to ethical obligations.
My advice for any lawyer exploring AI is to start small and without sensitive data. Create an agent that sends a fun, daily briefing summarizing your calendar, emails, industry updates – even scores from last night’s baseball game. The goal is building comfort and confidence through responsible experimentation. From there, the possibilities grow quickly.”
Experience some AI magic? Tell us about it at a.mills@newsweek.com
Tap or click here to get this newsletter delivered to your inbox.
