It wasn’t too long ago that headlines about artificial intelligence (AI) were mostly concerned with splashy news stories about game-winning models like Google’s AlphaGo or IBM’s Watson.¹ That’s all changed today as the breakneck pace of advancements in AI technologies have pushed adoption to the top of CEOs’ agendas.

“AI and genAI will empower companies to realise significant gains in productivity and efficiency, while simultaneously opening opportunities for new value streams, ecosystems and products.”

- Patrick Winter, EY Asia-Pacific Area Managing Partner

According to EY’s October 2023 CEO Outlook Pulse, 43% of leaders are already invested in AI, while another 45% plan to in 2024.² This increase in popularity has been accompanied by sharp increases in AI investment, with nearly 50% of technology officers viewing these as strategic priorities. ³ Off-the-shelf and custom AI spend globally is expected to grow to US$64 billion in 2025, growing 50% faster than the overall software market.⁴

CEOs’ growing recognition about the need to act quickly on AI has only deepened in light of the runaway success of generative AI (genAI). GenAI’s large language models (LLMs) have kickstarted a revolution for the technology and are projected to potentially generate US$4.4 trillion in corporate profits annually. ⁵

“As the technology advances at speed, what is certain is that AI is set to significantly reshape the global economy,” says Patrick Winter, EY Asia-Pacific area managing partner. “AI and genAI will empower companies to realise significant gains in productivity and efficiency, while simultaneously opening opportunities for new value streams, ecosystems and products.”

The technology will also be key to furthering digital transformation, as companies work to rapidly build related systems that AI depends on, like large-scale data collection and management systems, cloud computing and scalable tech stacks. These transformations will happen across the organisation, adds Mr Winter. “AI applications can, and will, impact each element of the business, touching the whole value chain. Organisations, regardless of size, will need to rethink their strategy and operations by putting humans and machines together,” he says.

With a wide-ranging toolkit of capabilities, AI and genAI will be capable of unlocking new realms of innovation and growth for companies, primarily by enhancing an organisation’s operational efficiency through automation of repetitive and time-consuming business activities. By automating tasks like drafting emails or customer communications that dominate employees’ working time, genAI frees up previous “time and headspace to execute higher-level cognitive tasks that contribute to solving complex challenges”, says Mr Winter.⁶

GenAI can also augment a workforce’s productivity. Across several studies, AI usage has been shown to result in 66% productivity uplift, with the biggest benefits occurring in cognitively-demanding tasks. ^{7 8} It does this by providing more accurate, real-time data insights for better decision-making, says Cameron Wall, AI & data managing partner at EY Asia-Pacific.

GenAI could even make human-centric internal functions such as human resources (HR) more efficient, says Mr Wall. For example, EY’s EY.ai Workforce—developed in collaboration with IBM—leverages organisations’ domain HR knowledge to automate common processes such as drafting job descriptions and extracting payroll reports. ⁹

“Companies’ efforts are being stymied by internal resistance and the ‘grief cycles’ that naturally occur with new technologies when employees struggle to accept changes in operations”

- Cameron Wall, AI & Data Managing Partner at EY Asia-Pacific

This has made it an interesting year for gen AI, especially in terms of enterprise adoption, says Sooyeon Kim, AI lead and consulting partner at EY. “Unlike previous technologies and AI, which are mostly centred around unstructured data, gen AI is able to address niche areas once thought to be resistant to automation,” she adds.

The rapid advancement in the intelligence of current AI models is evolving its use cases beyond its natural language capabilities. “Before gen AI, the skillsets of AI were impressive but narrow –recognising objects in images, quantifying language in text, or crunching numbers in massive data sets,” says Mr Winter. But LLMs can now do things previously thought uniquely human, like creating art, writing books and making music, he explains. “There are emergent, unexpected capabilities for AI beyond LLMs across areas such as pharmaceuticals, education, materials science and manufacturing, that have yet to be programmed,” he says.

In drug development, AI is already being used to greatly accelerate the pace of research and drug or vaccine discovery.^{10 11} Ms Kim also highlights the growing role of AI in forecasting decision-making for procurement by utilising large amounts of data such as commodity futures or currency exchange rates.

These are still early days, but the true value of the technology lies in its ability to execute tasks at a scale beyond human capacity, without the added costs of additional headcounts. Adopting AI can also imbue companies with “an edge on their peers”, says Mr Winter; conversely, failure to move quickly on the technology could damage their long-term competitiveness.

“Businesses need to start thinking about how the intelligence of AI could transform their business and ask questions such as, where the opportunities are for value creation and augmenting the potential of the workforce.”

Despite the optimism, AI adoption is lagging. Companies’ efforts are being stymied by internal resistance and the “grief cycles” that naturally occur with new technologies when employees struggle to accept changes in operations, says Mr Wall. Many companies are also not looking beyond AI’s cost-cutting potential to see how it can drive transformative change. According to Mr Winter, this can be attributed to issues including the lack of specialised talent, cybersecurity concerns, and misalignment between tech and business. In his view, the biggest blocker to deployment however, is “the lack of confidence”, both internally and externally.

“Much of the uncertainty comes from the evolving state of global regulation, especially as concerns grow around data privacy, misinformation and intellectual property rights around AI,” he explains. “Our own research shows that 65% of CEOs believe that more work is needed to address AI’s social, ethical, and criminal risks, but technology is evolving faster than governments can keep up.”

65%

of CEOs believe that more work is needed to address AI’s social, ethical, and criminal risks

Click to see CEO Outlook report

Building confidence is thus key to ensuring success, but doing so will require companies to take more responsibility over communicating how they are approaching AI, and consciously embed responsible AI guidelines and frameworks in their business. To address this, EY developed its EY.ai Confidence Index to provide a comprehensive information platform that can bridge gaps between stakeholders.

Another hurdle in the way of AI adoption is the sequestering of expertise within specific functions. Realising AI’s exponential value will however require companies to move away from silos towards a holistic, cross-functional approach that prioritises collaboration.

For example, one EY banking client’s siloed approach to improving its debt collection yielded only modest results. To resolve the issue, Mr Winter says the EY team developed an AI-based solution that cut across its risk, strategy and operations teams, and leveraged the full potential of the bank’s people, processes and technology. This resulted in a US$25m reduction of credit losses annually.

“Reshaping an organisation from end-to-end can optimise performance, enrich customer and employee experiences, and ultimately tap into new sources of sustainable growth,” he says.

“Many leading companies are realising that AI adoption is not about sourcing specific AI technology, but building an entire pipeline to collect data, train and retrain algorithms, monitor them, and so on.”

- Sooyeon Kim, AI Lead and Consulting Partner at EY

Ms Kim adds: “Many leading companies are realising that AI adoption is not about sourcing specific AI technology, but building an entire pipeline to collect data, train and retrain algorithms, monitor them, and so on.”

A common theme among critics of AI is its impact on the workforce, but the upticks in innovation could actually result in more jobs and demand for new skills in unexplored areas.¹² There is however a need for training and supporting employees as they undergo this transition.

“We know that changes to how we work are not always comfortable, even when they are positive, which is why we’re supporting our clients to leverage AI in the workforce to create that seamless human-AI collaboration,” says Mr Winter. To address this issue internally, EY and Microsoft developed EY Tax CoPilot, an education and enablement programme that would equip EY Tax professionals with the skills to leverage AI to unlock more productivity.¹³

Mr Wall suggests that leaders also consider how they can train employees about the limitations, realities, and potential of AI to limit risks such as AI hallucinations and prepare their workforce “to get through the grief cycle of any organisational change”.