How GenAI and complexity challenge assumptions and business models

What gets lost in the pursuit of use cases and efficiency is that GenAI is not just an efficiency-enhancing technology like cloud computing. It is more like the birth of the personal computer or the World Wide Web — a foundational, transformative technology. GenAI is increasing the speed and scale of disruption, encompassing segments of business models previously considered off-limits, and requiring companies to fundamentally rethink the way they operate and innovate those models.

The insights in this article are based on several sources, including multiple sessions at the 2024 Innovation Realized summit, two facilitated workshops with close to 50 clients exploring the impact of AI on business models, and a series of interviews with EY and external subject matter experts.

Definitions

• Operating model: how a company is structured internally to conduct its core activities, including its talent, workflows, processes, and information systems.
• Business model: how a company creates, delivers and captures value in the market. Business models and operating models as defined here are distinct but interrelated concepts since a company’s operating model helps shape the value creation component of its business model.
• Business model innovation: the process of identifying, developing and commercializing new business models to retain competitive advantage.
• Value proposition: the key benefit(s) customers gain from choosing a company’s product or service. At the core of any business model is a value proposition (e.g., convenience, price, and/or customizability).
• Value chain: the sequence of processes and activities used by a company to create and deliver products and services to customers.
• Value driver: any individual step, activity or process that adds value within a value chain. 

Second, the full potential for business model and operating model disruption comes from the combination of not just GenAI and traditional AI, but also other disruptive technologies and platforms. These range from technologies already in market, such as mobile apps and the Internet of Things (IoT), to ones whose full impact will be realized in the years ahead, such as blockchain. Also included in the list of potential disruptors are new and emergent capabilities that AI may gain in the months and years ahead.

The 2024 EY Reimagining Industry Futures Study finds that companies are investing in a broad range of technologies, with considerable variation across sectors.

Which of the following technologies is your organization investing in?

The current adoption of technologies varies significantly, with GenAI being the most nascent, followed by AR or VR, and blockchain.

What is the deployment status of emerging technologies that you are currently investing in?

Yet, the same survey also showed that 75% of enterprises do not adequately understand how emerging technologies can be combined to create value. This is a critical shortcoming, since combining technologies enables benefits that would not have emerged from any single technology — disruption is more than the sum of the parts. For instance, the combination of GenAI and robotics could deliver not just robots that can go far beyond human capabilities — lifting more, moving faster, with greater resistance to injury and less need for downtime — but also robots endowed with the world’s knowledge base and insight as a core capability. 

Several companies are actively working on combining AI and robotics. NVIDIA reports that more than 1.2 million developers and 10,000 customers and partners are using its Isaac robotic operating system and Jetson partner ecosystem to develop AI-powered robots.

GenAI can endow robots with more robust capabilities, including increased autonomy and the ability to learn in adaptive, unsupervised ways. In August 2024, Figure, a California-based startup, launched its Figure 02 robot, with GenAI-enabled capabilities that include speech-to-speech reasoning and conversation with humans through onboard microphones and speakers connected to custom GenAI models, as well as fast common-sense visual reasoning from robot cameras connected to an onboard vision language model.

GenAI’s ability to improve and develop new, emergent, capabilities at a rapid pace, along with a general acceleration in the pace of technology adoption since the pandemic, also means that leaders should not underestimate the pace of change. GenAI’s emergence took the business world by surprise. Parts of a value proposition or business model that seem disruption-resistant today may not be off-limits for long.

These two attributes — the combinatorial power of multiple emerging technologies and the disruptive potential of the full universe of use cases to which they could be deployed — have the potential to reshape operating and business models in fundamental ways.

Reshaping operating models

AI is significantly impacting companies’ operating models — the internal constructs that define how a company operates, including its workflows, processes, talent structure and information systems. Since the use cases businesses are exploring largely seek out efficiency and productivity gains, they inevitably impact talent structures and work processes.

Initially, many firms may be looking at how best to apply AI to existing business processes, but the real potential lies in reimagining processes from the ground up, so they are optimized for an AI-driven world.

Over time, this disruption — especially in concert with the next generation of emerging technologies — will reshape not just business processes, but entire business functions and whole operating models. A prime example is the emergence of autonomous AI agents, a market that some independent estimates see growing by a compound annual growth rate (CAGR) of 42.8% between 2023 and 2030. Gartner predicts that by 2028, one-third of interactions with GenAI services will use action models and autonomous agents for task completion.

“The next phase of AI capabilities will enable AI-based agents,” says Matt Barrington, EY Americas Emerging Technologies Leader. “Agents and humans will work collaboratively, with humans making the decisions. Over time, agents will gain greater autonomy and ultimately, a swarm of physical and virtual agents may even be able to run your business. At that point, we will see a ‘digital flip,’ in which the winners move to entirely different ways of running and operating their businesses, with AI doing much of the heavy lifting. The real tipping point will be seen with multiple functions moving to an agent-led model, which will open up entirely different ways to create value.”

The extent to which companies and regulators will actually permit technologies to operate in highly autonomous ways is an open question. Certainly, there are many applications in which the risk of such autonomy outweighs the potential benefits. But, even without getting all the way to completely autonomous business processes and organizations, these technologies have the potential to reshape companies’ operating models in fundamental ways. 

The biggest reason for this is that GenAI has taken automation in directions that caught most of the business world by surprise. While many companies were working with traditional, rules-based AI to automate parts of their operations, the conventional wisdom was that AI was unlikely to be ready for creative and higher-order knowledge work anytime soon. It was generally accepted that soft skills such as empathy, communication and nuance would remain resistant to automation.

Those assumptions changed in a flash with the launch of ChatGPT. GenAI models proved adept at taking on  work that had previously been considered off-limits, from writing articles and code, to creating images and video. They have also gained increasing proficiency in soft skills such as communication and displaying (or rather, simulating) empathy. 

We live in a time when everyone expects disruption — but the disruption we get may not be the disruption we were expecting. This reality — driven by GenAI’s ability to take automation in new directions and acquire new capabilities — means that, for many companies, technology will now be able to comprehensively disrupt the entire value chain and business model, including parts previously considered safe from disruption. Other technologies may further challenge assumptions about the nature, speed and extent of disruption.

Some of the biggest impacts may be in sectors that have experienced relatively little disruption so far. A prime example of such a sector is automotive.

While developments such as ridesharing and electric vehicles have certainly started to shake things up in recent years, the extent of disruption in the automotive industry is dwarfed by that seen in scores of other sectors, from retail to media and entertainment. Automotive is  a prime example of a sector that is being disrupted by a confluence of technologies and tech-enabled platforms, including not just AI, but also electric vehicles (EVs), autonomous vehicles (AVs), ridesharing platforms, and the IoT. Some of these are well developed and mature (e.g., ridesharing platforms), while others are nascent (e.g., AVs), but considering their combined impact allows us to envision how legacy business models could be fundamentally reshaped in ways that would not have been possible previously.

For an illustrative company in the legacy automotive sector, its core competencies — and biggest value drivers — have been in activities such as:

• Vehicle design
• Engineering and core software
• Manufacturing and managing a global supply chain
• Sales and marketing, including through the management of a dealer network

The combination of technologies listed above has the potential to automate, commoditize and diminish the value of these legacy core competencies. The application of GenAI to everything from R&D to design is bringing greater automation to these functions and diminishing their significance as value drivers. In addition, engineering could become less critical with the growth of EVs, which are mechanically simpler in many ways, making engineering-related attributes less valuable as competitive differentiators. These include key aspects of performance (all EVs are capable of instantly delivering maximum torque) and reliability (EVs have fewer moving parts that are subject to wear and tear, while regenerative brakes are more efficient and longer-lasting than conventional ones). The mechanical simplicity of EVs might similarly diminish the significance of manufacturing as a value driver. Vehicle design and branding may become less of a differentiator in a world of AVs and ridesharing, since passengers typically care less about many design features than owners. Sales channels could also be disrupted; some EV companies are already bypassing traditional dealership networks and selling direct-to-consumer.

There are still many unanswered questions about this scenario. EVs remove mechanical complexity but add electrical and battery complexity — will the net impact yield simplification? How will technological innovation affect the cost of batteries, as well as the long-term reliability, life span and depreciation of automobiles? Will dealer networks survive or be disrupted?

While these questions will only be answered with time, it’s not hard to envision a scenario in which the combined impact of the changes listed above sees automobiles going the way of the personal computer — commoditized hardware platforms in which value migrates to software and services. In an early indication of such a shift, Foxconn, the world’s largest contract manufacturer of electronics, has entered EV manufacturing in a big way, as have a spate of Chinese tech giants, signaling that the future of automotive manufacturing may have more in common with electronics than internal combustion engines.

With their traditional value propositions and value drivers diminished, automotive companies would need to identify new value propositions and building new business models — often in concert with an ecosystem of external partners. Examples include charging infrastructure. In 2024, a consortium of seven auto manufacturers joined forces to build charging stations across the US. Such infrastructure could enable new value pools, not just by generating fees, but also potentially by creating opportunities to monetize user data. With access to user data, and as cars become increasingly autonomous, in-vehicle experiences could become another new value pool, and the combination GenAI and IoT could enable new experiences that are highly personalized and context-specific.

EY analysis has identified a large number of future value pools for the automotive sector, and highlights three “value megapools” that collectively represent a revenue opportunity of more than US$660 billion by 2030:

• Supercharge the future: batteries and charging, including EV battery production, active materials and components, EV public charging solutions, and energy storage systems.
• Redefine the vehicle architecture: shifting the automotive from hardware-defined to software-defined, including enabling technologies, advanced driver-assistance system or autonomous vehicle components, data monetization, and software-based repair and maintenance.
• Close the loop: circular models aimed at reusing and recycling batteries and other materials.

AI is the most visible example of a larger shift. Companies are operating in an environment characterized by faster change, more interconnected impacts and greater uncertainty, driven by not just technological change, but also geopolitical volatility, demographic shifts and climate change.

This has implications not just for how business models are disrupted but also how companies approach business model innovation itself. Firms will need to rethink how they approach business model innovation to be ready not just for AI but also other changes that lie over the horizon.

Here are some ways in which businesses and leaders could start to rethink their approach to business model innovation:

2. Pursue a portfolio approach to business model innovation

In an uncertain operating environment, companies thrive by exploring and developing multiple business models simultaneously, which provides more flexibility in changing circumstances.

This idea has been used to greatest effect in the technology sector, where many large companies have thrived by innovating multiple new business models, one example being Uber’s development of new business models in food delivery, scooter rentals and freight logistics. Meanwhile, the corporate venture arms of many large companies, such as Google, also make multiple simultaneous investments in disruptive business offerings.

As with any investment strategy, this should ideally entail a considered approach to capital allocation across the portfolio. More than a decade ago, an article in the Harvard Business Review laid out the golden ratio of investing in innovation.¹ The authors found that “companies that allocated about 70% of their innovation activity to core initiatives, 20% to adjacent ones, and 10% to transformational ones outperformed their peers, typically realizing a P/E premium of 10% to 20%.” Interestingly, they found the performance of investments in the portfolio was roughly the inverse of this ideal allocation. Over the long term, core initiatives (those optimizing existing products for existing customers) accounted for 10% of returns, adjacent initiatives (those expanding from the existing business to “new to the company” businesses) accounted for 20% of returns, while transformational initiatives (those developing breakthroughs and inventing things that don’t yet exist) accounted for 70% of returns. Investments in transformational and disruptive opportunities may be a small proportion of overall resource allocation, but they can deliver outsize returns over time, as legacy models are subject to disruption.

While these best practices were published over a decade ago, they are far from being widely adopted across the business world. Today, they are more salient than ever, and business leaders should develop a portfolio approach to their investments in GenAI and other technologies as they navigate an environment of rapid change and increased uncertainty.
 

3. Make business model innovation an ongoing process

The days of gathering senior leaders every few years to develop a new five-year strategic plan are numbered. In a world of faster change, business model innovation needs at minimum to be conducted every quarter. Even better, it should become a continuous process, with companies developing the requisite capabilities and competencies to do this on an ongoing basis.

Developing business model innovation as an ongoing capability is particularly important as companies look to adopt a portfolio approach to business model innovation. EY teams recently worked with a major convenience store chain to help them develop a future vision for the convenience store of 2030. A key feature of their future vision was the digital storefront, enabled by tablet interfaces. However, it soon became apparent that executing this vision would be challenging, since many store employees are not native English speakers and have low education levels. So, they revised their future vision to one empowered by auditory interfaces and information flows instead of text-based tablets. The company is interested in exploring additional future states and business models, and going through the process of pressure-testing and iteratively refining this business model demonstrated to them the value of being able to do this repeatedly and at scale. So, EY teams built them a turnkey capability to conduct business model innovation on an ongoing basis.

Another company looking to make business model innovation an ongoing process is Volkswagen Group, which has launched an “AI Lab” focused on developing disruptive offerings powered by AI. The AI Lab serves as a global hub for innovation, identifying new product ideas and collaborating with multiple technology partners to develop and explore a range of disruptive offerings.

President Eisenhower famously said that “plans are useless, but planning is indispensable.” A variant of this observation is applicable here. In a more complex and uncertain world, the goal is not so much to define a single specific destination and build the roadmap to get there, as much as to identify scenarios and directions of change, and build capabilities that will be needed across multiple potential scenarios within this general direction of change.

Summary

For all the interest and activity around GenAI, most companies have not yet focused on its ability to disrupt their business and operating models. This is a critical gap, because GenAI has the potential to reshape existing models at much greater speed and scale. To understand its impact on their business and operating models, companies should consider the combined impact of every potential use case and multiple technologies, and should challenge existing best practices for business model innovation.