Skip to main content Skip to footer
Pages

Executive summary

Before generative AI captured the popular imagination in late 2022, the ability to create new things—a competitive analysis, business presentation or piece of software code—was seen as an exclusively human trait. Now, with the showstopping debut of Open AI’s ChatGPT, anyone with a computer can witness generative AI systems respond to a prompt with new content and ideas at mind-bending speed.

At first blush, this could seem threatening. But as time goes on, the promise of gen AI will grow increasingly clear: Rather than replace humans, this technology will enhance and augment human intelligence and decision-making, making us better at what we already do.

Businesses already can integrate gen AI tools, safely and responsibly, into their workflows. But as gen AI further permeates enterprise technology stacks, it will expand beyond simply automating single tasks. Multiple gen AI agents will collaborate with each other to orchestrate all the processes, systems and pools of knowledge needed to execute a complex series of interconnected tasks, from modifying a product design, to figuring out your PTO based on your upcoming workload. And rather than maneuvering through disparate systems, apps and data, workers will use a single interactive and conversational interface that makes all the necessary connections.

In short, gen AI will change how we work by acting behind the scenes to pull together all aspects of the business and provide a unified access point for consumers and employees alike. As gen AI matures, it will make our current levels of productivity seem quaint, while changing—at the deepest level—the way businesses innovate, make decisions and organize themselves.

The risks of gen AI are well-documented. Many businesses are hesitant about incurring a major security or ethics breach—not unlike the early days of PCs, the internet and mobile computing. But like those technologies, gen AI will move through its current era of vast disruption to become an unquestioned part of the fabric of work. With due diligence, governance and a phased implementation, these new tools can, and should, be safely deployed without constraining the potential gains in innovation, efficiency and productivity.

Leaders who understand the scale of what’s unfolding and begin deploying gen AI safely today stand to gain more than the usual “first-mover advantage.” With gen AI, the potential gains for leading the way in its adoption are as limitless as the possibilities of gen AI itself.

Gen AI is already at work:

From writing FAQs to code, generative AI is already boosting productivity.

Know more
Image
Next up-orchestration:

Even more transformative is what will happen when gen AI agents start to connect and talk to each other.

Know more
Image
A spark for innovation:

Gen AI can supercharge humans’ ability not only to make and create, but to think.

Know more
Image
A decision making assistant:

With gen AI embedded in business systems and paired with the right data, decision making will be faster and sharper.

Know more
Image
A catalyst for better organizational structures:

To embrace gen AI, businesses will need to reshape organizational structures.

Know more
Image
Risks and aversions:

Understanding gen AI’s potential issues is key to successfully navigating this new frontier.

Know more
Image
How to start:

From starting small to choosing the right model, five tips for deployment success.

Know more
Image
The road ahead:

Gen AI delivers on its promise to improve the way businesses operate and humans work

Know more
Image

Gen AI: a breakthrough in automation

When the world was introduced to gen AI through ChatGPT, what stood out initially was its fluid chat interface that users could talk to naturally. The model would reply with seemingly thoughtful answers, scrolling out across the screen as if typed by an excited friend.

But it was what gen AI could do and make from those fluid chats that soon became the bigger story. As distinct from “traditional AI” systems, which react to inputs by following pre-set rules, gen AI models can create information. If traditional AI in a streaming service can recommend a movie you might like, generative AI can, in seconds, write an original movie script precisely tailored to your individual tastes and requests. Or a symphony. Or software code.

In our September 2023 survey of senior business and technology decision makers at large businesses in the US and UK, in fact, 61% of executives expect generative AI to result in a complete business transformation. Already, this new technology is hard at work in diverse ways:

Customer service

Some companies are using gen AI to mine their knowledge management systems for meaningful data, which they then use to create new FAQs and guides. If the FAQ doesn’t solve a customer’s problem, gen AI can comb databases for technical info, and even suggest solutions based on similar cases. This frees up human agents to work on more complex customer queries, resulting in a higher level of service. In our survey, 50% of senior leaders said they were already using generative AI to pull data from customer conversations to address customer needs. 

Image
Wealth management

At Morgan Stanley,  wealth advisors have begun using gen AI to help make investment recommendations and ask general business or process questions, based on its enormous repository of research and structured and unstructured data, including text and video.

Image
Clinical trials

Pharmaceutical companies are using gen AI to help identify potentially overlooked insights in historical clinical trial data. This could shorten drug-discovery timelines, a breakthrough with major financial implications and even larger potential impacts on human health and longevity.

Image
Code creation and SDLC

Generative AI can significantly shorten the software development lifecycle. Paired with repositories of code, user requirements and testing scenarios, gen AI can create code snippets, construct user stories that align with business requirements and devise test cases that cover an array of scenarios. It can also create synthetic data, which enables robust testing without relying on sensitive data. In our survey, 61% of execs cited software development productivity as the area where gen AI could play the largest role in their workplace.

Image

The productivity gains from gen AI can be particularly dramatic among new or entry-level staff, who can quickly develop expertise that would otherwise take months of experience. This was apparent in a recent Stanford and MIT study, in which call-center reps who used gen AI were 14% more productive on average than those who didn’t. The gains were even greater among workers who had been on the job for less than a few months.

Even more transformative is what happens when gen AI agents start to connect and talk to each other. While early discussions revolved around using gen AI for specific tasks, such as coding or creating software, consider a future where those tasks are no longer necessary because, instead, a detailed prompt replaces the need to write extensive lines of code.

This new way of interacting with a digital system compels us to question whether traditional apps and websites will even be necessary in the future. As generative AI becomes more advanced, it could usher in an era where digital interaction is far more intuitive, immediate and tailored to individual needs, going beyond what traditional apps and websites can offer. The true revolution of generative AI is in opening doors to these previously unimagined possibilities.

From automation to orchestration

Consider that a big reason for the immense, near-overnight success of ChatGPT was its ability to act as a single interface, a “one-stop-shop,” to a vast wealth of human knowledge. That one-stop-shop model will continue as gen AI gains traction, but it will also expand. As multiple gen AI agents begin collaborating, they will act as an orchestration layer, wrapping enterprise systems, specialized data sets, resources and processes into one cohesive unit. For leaders and workers alike, the primary point of contact with the organization will be a conversational AI assistant that functions much like a human knowledge worker, except that it has instant, real-time access to the information and resources needed to do its job in the form of the dataset that’s feeding the generative AI tool.

A typical workflow might go as follows: you need to modify an existing product to enable features for people who have a visual impairment. You ask your gen AI agent to analyze the existing

product and suggest some alternatives. None of its suggestions hit the mark, so you provide it with more specifics of what you’re looking for. As the creative juices start to flow, and the product modifications take shape, the gen AI agent continues to help by refining the product, checking it against regulatory requirements, generating blueprints and prototypes and keeping other team members in the loop—all the way through marketing and promotion to launch day.

Early iterations of this concept are already in development, with AI-powered task management projects like AutoGPT and Baby AGI leading the way. While they are not without their flaws, these projects offer a glimpse into the future of gen AI assistance.

Once mature, this single point of contact model will be more than just a convenience or operational shift. It will completely change how we innovate, make decisions and structure our organizations.

 

Unbound innovation

Traditionally, productivity gains have flowed from process automation. With gen AI, the gains will also come from innovation, as this new technology supercharges humans’ ability not only to make and create, but to think. Leveraging large language models, for instance, gen AI assistants can summarize, distill and compare books and research papers, vastly increasing the supply of intellectual “raw material” that fuels new ideas—new ideas that can be turned into valuable realities by anyone able to articulate those ideas via prompts and conversation with their tireless, always-interested gen AI assistant.

When innovating today, we often have to deal with rigid systems. We must shape our ideas and products to fit the existing tech landscape, with its creaky legacy systems that often do not support groundbreaking innovation. In many cases, we have to focus more on what’s technically possible, on what we can do, rather than on what would truly fulfill our objectives.

Gen AI blows through these many technology restrictions. It introduces a flexibility that hasn’t been seen before, as it’s not dependent on strict system requirements and can handle unstructured data. In this environment, productivity gains will flow from innovation, not just by doing the same things faster through automation.

The flexibility of gen AI is already evident. The CAMEL project, for instance, has multiple gen AI agents that adopt distinct personas to tackle a problem, as if you’d brought together a mini think tank that includes a microbiologist and a quantum physicist to brainstorm a solution.  

Ultimately, gen AI promises to be a catalyst for human creativity, not a replacement for it. By effectively extracting and utilizing AI-generated insights, individuals can refine and implement strategies that maximize their own innovative thinking, propelling the boundaries of human creativity.

 

From innovation to impact:
a new level of strategic decision-making

Innovative thinking opens up vast horizons; decision-making determines the outcome. Once an organization incorporates generative AI into its business systems and establishes a unified backend for data access, it can significantly speed up and sharpen strategic decision-making processes.

More data driven insights: Generative AI will enable access to a steady, reliable pool of data while also adding layers of depth and usability. LLM-based data augmentation will enrich business datasets, making their AI models even smarter. Gen AI can process both structured and unstructured data, as well as legacy data, where it can identify valuable data points that were previously overlooked. This transforms data from just being available to being deeply accessible and insightful.

Consistency is crucial: With generative AI as the interface to business systems, everyone in the organization is working from the same information—gone are the disparities in data sets or discrepancies between departments. This consistent use of data greatly improves organizational performance and reduces the likelihood of errors or tampering, resulting in streamlined and efficient processes.

Differentiation through decision-making: Leaders seeking competitive differentiation will embrace generative AI to identify and meet their most significant key performance indicators (KPIs). By identifying the KPIs that set them apart—considering workflows, policies, legacy data and associated analytics—leaders can use generative AI to help them make decisions that help them meet these goals.

New organizational structures

Generative AI might fundamentally reshape the inflexible, department-based organizational structures that have existed for nearly a hundred years. Just as electrification and industrialization forced us to reimagine how businesses are structured and operate, an AI assistant that can take on and connect many of the tasks of functions like marketing, legal, procurement, operations, R&D and sales will spur businesses to reconsider whether traditional, siloed configurations still work.

This point isn’t about needing fewer employees; it’s about reimagining how our current teams operate. By tapping into previously disconnected workflows, applications and knowledge bases with AI assistants, teams can stop working in silos and collaborate to reach goals and make meaningful contributions. Rather than disappearing, jobs will become outcome-focused and reliant on AI to access skills and knowledge.  

Generative AI will also shake up how we think about organizational roles, as some technical skills become less necessary and other, more specialized capabilities grow in importance. Consider cybersecurity. As bad actors get creative with AI prompts to extract sensitive information, businesses will need professionals who can anticipate these attempts at deception. This means roles that once focused heavily on the technical side of security now require a high degree of creativity and innovative thought.

Further, as gen AI becomes increasingly adept at problem solving, it will be up to human workers to get better at problem finding, as they will be the ones to prompt gen AI to find innovative issue resolutions and opportunities. A new diversity of skills will be needed, including an understanding of human nature (sociology, psychology, anthropology), process design and optimization (design thinking, Six Sigma, industryspecific knowledge) and audience engagement, both intellectually and emotionally, through storytelling and design.

Moreover, collaboration—between employees with varying skills and between employees and technology—will be pivotal to effectively harness this diverse knowledge. Businesses will need to experiment with flatter organizational structures and devise flexible frameworks that encourage and reward collaboration.

Risks

As with any emerging technology, generative AI is not without its pitfalls. Understanding these potential issues is key to successfully navigating this new frontier.

Inaccuracies and "hallucinations"

Generative AI relies on the data it’s been fed to make predictions and generate outputs. However, it sometimes creates outputs that are inaccurate or completely fabricated—termed “hallucinations.” These hallucinations could lead to misinformed decisions or actions, potentially causing significant issues for a business.

interacting
Bias and ethics

AI is only as unbiased as the data it’s been trained on. If that data contains biases, the AI can, and will, replicate and amplify those biases in its outputs. Ethical issues can also arise around privacy, a lack of consent or agreement on the use of copyrighted data used for training, and misuse of generated content, all of which businesses need to consider.

interacting
Redundancies

The adoption of generative AI might lead to some job roles becoming redundant, particularly those involving repetitive or data-heavy tasks. While this could lead to increased efficiency, it also brings up questions around job displacement and the need for re-skilling. It’s important to remember, though, that new tech also creates new roles and opportunities that previously didn’t exist, thereby contributing to more per capita income, more prosperity and more upward social mobility.

A recent study corroborates the idea that AI has not resulted in job loss. Economists at the National Bureau of Economic Research found a 5% increase in the number of openings for highly skilled jobs that had been considered vulnerable to AI, such as white-collar office work. The timeframe for the study was 2011 to 2019, the period when businesses started using deep learning to automate tasks. The researchers concluded that new technology can increase demand for more skilled workers even when it replaces those who do routine work.

Sam Altman, the CEO of OpenAI, recently explained that while gen AI today is good at doing “parts” of jobs, it’s not very good at all at doing “whole” jobs. In the short and medium term, if not beyond, there will always be a human in the loop.

interacting
Security

As with any digital tool, generative AI systems are vulnerable to cyber threats. As we’ve pointed out earlier, there’s a risk they could disclose sensitive information. The necessity for a robust cybersecurity protocol is clear.

Another threat is prompt injection, a technique that coaxes AI models to give away information they shouldn’t and doesn’t always require advanced technical skills to carry out.

Security risks, privacy risks and reputational harm were the top three concerns of execs in our survey with generative AI. It’s essential for chief security officers to fully grasp all the ways generative AI could be compromised. It’s only by understanding every avenue of attack that they can maintain robust defenses.

interacting
Regulation

The burgeoning and patchwork regulatory landscape around generative AI, with measures like the proposed EU AI Act, presents a significant factor that businesses need to understand and navigate. Nine in 10 execs in our survey said they’re struggling to understand the impact of generative AI on business regulations.

interacting
Sustainability

Generative AI has the potential to contribute positively to sustainability work, from aiding in regulatory reporting to analyzing data to create innovative solutions. Nearly all execs in our survey (96%) said generative AI would positively impact sustainability because it will democratize technology access. However, it’s important to acknowledge the negative impact of its energy and compute requirements. Similar to blockchain technology, the energy demands of generative AI could cause a backlash. 

interacting

How to start

It’s unwise to ignore the risks that come with generative AI. Instead of pretending these risks aren’t there, it’s crucial to establish the necessary safeguards to manage them effectively. This means understanding where caution is required and where safety and security measures need to be in place, all the while taking advantage of the technology’s potential. To effectively do this, you need to:

1. Start now

Early adopters of generative AI will have the advantage of better data management and building the skills necessary for successful use of the technology. Understanding your data and AI models takes time, and this time lost by late adopters can’t be bought back, meaning there’s a clear early mover advantage.

2. Start small

Identify areas in your organization where the potential impact of risks is lower. This way, you can simultaneously run new methods alongside existing processes. This helps you gradually learn about generative AI while minimizing potential disruptions. Your focus should be on creating an environment that encourages quick experimentation and innovation. The key is to cultivate a culture that’s not afraid to try new things.

3. Choose your models

Different gen AI models excel at different kinds of tasks, and they occupy different points on a spectrum of openness and risk. It’s important to determine what your organization needs to use in order to provide the most value and mitigate the particular risks to your business. Some models are publicly available, yet the data that has been used to create them is opaque, while others are smaller and more niche, based on a particular industry or subject matter and the data can be fully analyzed and understood. Making these choices will inform the risks that are present.

Public models

There will be two to five large, “general” public models, trained on public, limited private and sometimes indiscriminate data.

Expert models

Organizations will build models themselves, using private IP, to serve a specific purpose or niche.

Industry models

Big players like Microsoft, Google, AWS and OpenAI will offer customizable, fit-for-purpose models as-a-service.

4. Prioritize traceability

It’s essential to understand the data you’re using to train your gen AI models. This includes knowing the quality of the data and any potential risks it might expose internally or externally.

5. Monitor gen AI with gen AI

Generative AI systems themselves can be used to critique and assess potential risks. For example, you could use them to check if a new output has any ethical biases, or if there’s a reputational risk associated with publishing a piece of content.

The road ahead

Ernest Hemingway’s oft-quoted insight that bankruptcy happens in “two ways: gradually, then suddenly” is also how history will likely remember the rise of generative AI. ChatGPT’s splashy debut last November was the “suddenly” part. In the space of mere days and weeks, the non-tech, nonbusiness world underwent a crash course in gen AI, mastering new terminology— “large language model,” “training set”—and speculating, often wildly, about the impact of this new technology on human affairs and society. The lone point of agreement between the utopians and catastrophists was on the violent speed and scale of change: that gen AI, for worse or for better, represents at least an “inflection point,” if not a full “disruption,” in the hitherto smooth unfolding of human history.

Lost in the breathless noise we find the “gradually” part. The development of gen AI will likely be an incremental process consisting of decades of continual refinement and improvement. Gen AI’s impact on the world of business and the changes described above will be dramatic and fast-moving, headed in a familiar direction: towards greater efficiency, greater productivity and—most importantly—greater harmony. Yes, as described above, gen AI will break down the walls between departments and functions within a business, helping them work in concert towards a common goal, but this institutional harmony is downstream of an improved and more harmonious relationship between individual human beings and the world around them, a world that gen AI makes much easier to control and navigate.

 

The job of leaders in these coming years will be to keep that harmony in mind and use it as a North Star in making the choices and taking the risks that gen AI already presents. The excitement around this new technology flows from its potential to forge a new, nearly frictionless relationship between human beings and machines. What may matter more, ultimately, when we look back on these turbulent years, is how gen AI helps human beings work with other human beings.

About the authors

Duncan Roberts
Cognizant Research

Currently with Cognizant Research, Duncan Roberts 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 and launching new innovative products. He holds a Masters in Philosophy and Classics from the University of St. Andrews

Email: Duncan.Roberts@cognizant.com
Linkedin: https://linkedin.com/in/duncan-roberts-16586022/

Naveen Sharma
Vice President and Global Practice Head, AI & Analytics, Cognizant

Naveen Sharma is Vice President of Cognizant’s AI & Analytics business. He is an accomplished technology services executive who excels in blending strategic vision with tactical execution to achieve business agendas. He is focused on driving growth through thought leadership, innovation, pre-sales, offering development and portfolio management in this space.

Naveen has over 25 years of experience in services and technology, working with premier brands such as BMS, Sapient and IQVIA and has experience in consulting with Fortune 500 firms across their data and analytics journey.

At Cognizant, he has fulfilled several roles, including practice head for its Enterprise Data Management business, AI&A practice leader for life sciences, and service line sales leader for life sciences.

Email: Naveen.Sharma@cognizant.com
Linkedin: https://www.linkedin.com/in/4naveen/

Babak Hodjat
Vice President of Evolutionary AI, Cognizant

Babak Hodjat is Vice President of Evolutionary AI at Cognizant, and former co-founder and CEO of Sentient. He is responsible for the core technology behind the world’s largest distributed artificial intelligence system. Babak was also the founder of the world's first AI-driven hedge fund, Sentient Investment Management. He is a serial entrepreneur, having started a number of Silicon Valley companies as main inventor and technologist.

Prior to co-founding Sentient, Babak was senior director of engineering at Sybase iAnywhere, where he led mobile solutions engineering. He was also co-founder, CTO and board member of Dejima Inc. Babak is the primary inventor of Dejima’s patented, agent-oriented technology applied to intelligent interfaces for mobile and enterprise computing – the technology behind Apple’s Siri.

A published scholar in the fields of artificial life, agent-oriented software engineering and distributed artificial intelligence, Babak has 31 granted or pending patents to his name. He is an expert in numerous fields of AI, including natural language processing, machine learning, genetic algorithms and distributed AI and has founded multiple companies in these areas. Babak holds a Ph.D. in machine intelligence from Kyushu University, in Fukuoka, Japan.

Email: Babak.Hodjat@cognizant.com
Linkedin: https://linkedin.com/in/babakhodjat

Acknowledgments

The authors would like to thank Catrinel Bartolomeu, Director of Storytelling & Content at Cognizant, Mary Brandel, Editor, Mykola Hayvanovych, AVP of AI, Cognitive Computing & Data Science at Cognizant, and Matthew Smith, AVP & Conversational AI Practice Leader at Cognizant, for their contributions to this report.