Canada
Existing businesses can’t become AI natives themselves. But they do need to stay vigilant as these AI upstarts seize new market opportunities. By actively studying how AI-native businesses put AI into the core of their operations and technology, established companies can reap the benefits of thinking and acting like their newest competitors.
Remember Monster.com? Back in the late 1990s, the emergence of the online job site was a godsend for job seekers everywhere who happily gave up their newspaper classified ad subscriptions for the new, ink-free web platform.
Within a few years, however, along came a new set of disruptors—from Indeed to LinkedIn to Glassdoor—each of which offered progressively more innovative features.
Soon, however, we may be asking a different question: Remember online job boards?
With the emergence of generative AI, we may soon experience a whole new approach to finding a job—or getting a date, mapping a route or booking a hotel. It seems like just yesterday these activities morphed into a new normal. Now, once again, they’re up for disruption.
Imagine, for instance, an AI-powered career coach. Rather than searching for a job, candidates would tell it their skills and goals, even their ideal company culture. In return, beyond offering up matches, it would analyze hidden patterns in job postings, news articles and social data to uncover opportunities at companies that hadn’t even advertised yet. Better yet, it would generate tailored cover letters, conduct practice interviews, even suggest negotiation tactics based on the potential employer. Who would make this happen? Most likely an AI-native business.
While AI-native businesses as we’ve defined them don’t exist yet, they soon will. These AI-first upstarts will insert AI (and increasingly generative AI) into the fabric of everything they do. And because their whole mindset is centered on AI-driven capabilities, the technology will drive how they think and do things.
Unburdened by the constraints of legacy systems and entrenched approaches, AI-native businesses will see AI not as a tool to be bolted on, but as the fundamental building block of their operations.
This pure AI mindset will allow them to capitalize on generative AI’s strengths, from the tailored experiences customers crave, to blazingly fast internal processes, to business models that, to a traditional business’s eye, appear inside-out and upside-down.
Like the digital-native businesses before them (Uber, Netflix, Venmo), AI-native businesses will also change how consumers behave. Consumers will begin to choose products differently, seek information differently and assess options differently, ultimately changing the way in which they interact with the digital world.
And that’s why traditional businesses need to pay attention. As consumers change their behaviors, businesses will need to accommodate these new ways of selecting, engaging and transacting. These changes will also open up opportunities for AI-native businesses to infiltrate enterprise functions, both collaboratively and competitively:
To execute newly hatched business models, organizations will need new governance policies, organizational structures and operating processes to deal with massively accelerated, dynamic and autonomous operations. And—vitally—they’ll also need to rethink their technology infrastructure to power the new AI-driven capabilities they offer.
This is where thinking like an AI-native business comes in. Existing companies can’t become AI-native businesses—that’s logically impossible. But they can and should learn from how AI-native businesses operate. By actively studying these newcomers’ innovative approaches, understanding their technology stack and adapting their own AI strategies, established companies can start to think—and even act—like the disruptive businesses that will put AI at the heart of all they do.
Fast uptake in the next decade
Adoption of generative AI could move quickly in the decade ahead.
Figure 1
Chart data reflects the highest possible adoption rates, as reflected in our high adoption scenario.
*To ensure we captured the full adoption rate through 2032, we included 2033 data in the calculation for this figure.
Source: Oxford Economics and Cognizant
It’s in the 2026–2030 time period—the era of “confident adoption”—in which we believe real generative AI change will occur as 31% of businesses will adopt generative AI, according to our most bullish scenario. While that’s just two short years away, it’s an ample buffer zone for businesses to prepare, if they start now.
At this juncture, the AI regulatory environment—currently a patchwork—will likely come into focus, and the technology will move beyond task automation to become a core component of business transformation strategies. Imagine a world, for instance, where routine customer inquiries are handled almost entirely by AI agents: autonomous systems that perform specific tasks in orchestration with each other to fulfill a larger goal. These AI agents would provide personalized support around the clock. Sales teams, instead of relying on generic pitches, would use generative AI-powered lead generation and customer profiling tools, allowing for precisely targeted approaches. Even back-office functions like scheduling, reporting and data entry will increasingly be automated.
But beyond imagining this world, our study brings many of these changes to light. As part of our analysis, we assigned “exposure scores” to 1,000 jobs currently being done by the US workforce.
This score doesn’t reflect the percent of workers who will be out of a job or their chance of losing a job. Rather, it’s the maximum percent of job tasks that could theoretically be automated or assisted by generative AI by the year 2032, weighted by the relative importance of those tasks.
A look at our exposure scores reveals occupation groups that are due for the most change with the emergence of generative AI—and represent prime areas for AI natives to both enter new markets and emerge as lean businesses. When looking at the areas above, for instance, customer service representatives will see exposure scores of 63.7%, sales representatives 65.6% and office and administrative workers 85.9%, all by 2032.
This is why—starting in the next two years when confident adoption of generative AI begins—forward-thinking companies will need to have a robust technological infrastructure, a strategy in mind for redefining the business model and a culture that encourages continuous innovation. For that, we can turn to AI-native businesses, which can offer lessons on how and where to make these important shifts.
Precise problem solvers
Foundation architects
Industry disruptors
To deliver on these new capabilities, AI-native businesses will build a technology stack that, in many ways, is unrecognizable to businesses operating today. This is because—unlike traditional businesses that typically see new technologies as a way to do existing things better—AI-native businesses see technology, particularly generative AI, as a way to do things that have never been done before.
From an AI-native business perspective, the tech stack is not a static thing that’s occasionally tweaked or updated. It’s a living, continuously evolving entity, built for a world where agility and continuous innovation are key to a business’s survival.
By pulling back the curtain on what an AI-native tech stack might look like, traditional businesses can begin thinking about the changes they need to make to bring generative AI capabilities into their own workplace and, ultimately, their business and operating models.
For AI-native businesses, the user interface (UI) is a fundamental rethink of how users (both customers and employees) interact with the services the AI-native business provides. Beyond considering button placement and screen design, an AI-native UI will be driven by well-crafted prompts that initiate complex actions across multiple systems or generate rich, multifaceted outputs. The UI design will facilitate a new human-to-AI interaction style, ensuring users can effectively guide AI agents and retain visibility and control over the results.
Established businesses should begin experimenting with this paradigm shift in UI design. It’s less about clicking through menus and more about conversational interaction to achieve goals. To make this mindset shift, businesses can start by asking themselves:
AI-native businesses won’t think in terms of a single, monolithic AI model. Instead, they will strategically employ a suite of models, each specialized for a variety of tasks. Such tasks could include text, image and process generation, as well as structured data modeling, prediction and decision optimization, uncertainty modeling, explainability and code optimization.
A diffusion model, for instance, might create stunning product visuals, while a large language model (LLM) would generate tailored marketing messages. Additionally, traditional machine learning (ML) models will be integrated intelligently to address specific problems and leverage the unique advantages of different AI approaches.
Similarly, established businesses should avoid viewing an AI model as a one-size-fits-all solution. It will be crucial to understand the strengths of each model type and select the right tool for the job. Businesses should invest in the expertise needed to determine what purpose these models genuinely serve and how they can be best deployed.
To enter a multimodel world, businesses should consider the following:
AI-native businesses will thrive on information—the wider they cast their net, the more powerful they will become. That’s why AI-native businesses will prioritize access to both structured and unstructured data over concerns about data cleanliness. Once they’ve established clear consent and transparency on how data will be used, they might tap into vast public datasets, third-party resources and customer data.
Additionally, they’ll use knowledge graphs (a way of representing data, entities and how they interlink) and vector databases (which are optimized to work with LLMs for fast, easy searches and data retrieval) to grasp complex relationships between data points.
For established businesses, it will be crucial to leverage vast quantities of information, both structured and unstructured. Regardless of how they do this, ensuring transparency in data acquisition will be critical.
However, acquiring massive datasets can be challenging. This is where businesses can get creative. They can use techniques like synthetic data generation, where realistic but artificial data sets are created, often using generative AI itself. They could also use semi-supervised learning, where models learn from a combination of labeled and unlabeled data. By prioritizing access to diverse information and employing these techniques, businesses will empower their generative models to perform at their best.
To improve their data IQ, established businesses should consider:
Generative AI models can be extraordinarily demanding from a computational power perspective, especially during training and fine-tuning. AI-native businesses will understand that their infrastructure needs to be both scalable and cost-effective, often opting for a hybrid cloud approach. This allows them to ramp up resources for intensive tasks and scale down when usage ebbs, achieving an optimal balance between cost control and performance. Hardware acceleration, utilizing specialized graphics processing units (GPUs) or tensor processing units (TPUs), will be essential.
Established businesses will need to proactively strategize for this level of flexibility and agility. By considering infrastructure implications well before scaling AI projects, businesses can prevent cost overruns and ensure smooth operations.
To make this infrastructure shift, businesses should ask themselves:
For AI-native businesses, machine learning operations (MLOps) is the cornerstone of maintaining a competitive edge. It isn’t a luxury, but rather the framework that allows these businesses to stay agile and responsive. Robust MLOps pipelines enable experimentation, fast deployment of new and improved models, and comprehensive production monitoring. These pipelines will ensure that AI-powered workforces are constantly learning, automatically retraining as new data becomes available, and seamlessly updating to optimize performance.
Established businesses need to shift their thinking toward AI as a dynamic system rather than a one-time project. By investing in MLOps, they can automate processes and shorten the critical feedback loop between real-world results and model improvements. Doing so will ensure their AI systems are continuously evolving to match the dynamic market landscape.
To make the MLOps shift, businesses should consider:
Ultimately, AI-native businesses will instinctively grasp an entirely new way of developing, evaluating, understanding and deploying AI systems. They won’t approach AI solutions with the mindset of building rigid, deterministic machines designed for perfectly replicable results. Instead, they will view their AI agents as evolving collaborators with the potential to be far more than mere tools. (For more on this topic, see Cognizant VP of AI Research Risto Miikkulainen’s recently published article, “Generative AI: an AI paradigm shift in the making?”)
This approach will shape their strategies in fundamental ways. AI-native businesses will prioritize building systems capable of continuous learning, understanding that “training” is never truly over. And they will design interactions that leverage the power of prompts and iterative feedback, guiding their AI agents with the finesse one might use when working with a highly skilled human specialist.
Similarly, established businesses should look to truly harness the power of generative AI rather than shoehorning AI into existing processes or evaluating it through the lens of old metrics. Doing so will require technology teams to think in a completely new way: a relentless willingness to experiment, shift away from a need for total explainability and accept that trust in the AI agent will be earned through rigorous observation rather than upfront programming.
In addition to thinking like—and ultimately acting like—an AI-native business, established companies will want to partner with AI-native businesses. While a relatively smaller number of traditional businesses may opt to become disruptors by building out their AI capabilities on their own—or spinning off an AI-native division of their own business—a larger number will look to partnerships to bring the fresh perspective of the AI native in-house.
The great majority will likely take a hybrid approach: partnering with AI-native businesses, using APIs to bring new AI services into their own environment and discerning where they absolutely must build AI capabilities themselves.
In all cases, established businesses can harness AI to transform existing processes and build innovative new offerings—and find ways to integrate the strengths of both traditional and AI-native models to secure their success.
Duncan Roberts
Associate Director, Cognizant Research
Duncan Roberts is a Associate Director in Cognizant Research. He joined the company in 2019 as a digital strategy and transformation consultant in industries ranging from satellite communications to educational assessment. He has advised clients on utilizing technology to meet strategic objectives and discover the art of the possible through innovation.
Before Cognizant, Duncan worked for one of the largest publishing houses in Europe, playing a leading role in the digital publishing revolution, helping transform their operations end-to-end and launching new innovative products. He holds a Master’s in Philosophy and Classics from the University of St. Andrews.
The author would like to thank the following for their contributions to this report:
