GenAI: Bridging the gap between intent and adoption

By Vishy Narayanan and Scott McLiver

“Does the world need another article on GenAI?” This is the prompt we entered into ChatGPT, Gemini (formerly Bard) and our own large language model, ChatPwC. In summary, they agreed that to help readers avoid information fatigue, content needs to be well-researched, engaging, and offer something novel. Thankfully, our recent Global CEO Survey — Asia Pacific gives us extraordinary new insight into how GenAI is genuinely perceived by business leaders. 

Several questions and doubts were revealed — many surprising — giving us the opportunity to respond to real-world thoughts and issues. Crystal-balling on where GenAI may take us is not our concern here. Speaking to the challenges that business leaders are finding, why we think this may be, how we’ve chosen to approach GenAI and how we can tangibly help, is.

Global CEO Survey— Asia Pacific: what did it reveal?

This year's survey, the 27th by PwC, revealed a gap between intention and adoption of GenAI. On the one hand, 65% of Asia Pacific CEOs anticipate that GenAI will significantly enhance the efficiency of both their employees and their own time. 76% also anticipate the need for their workforce to acquire new skills in response to GenAI advancement (surpassing global CEOs by 7%). Despite this awareness, 41% of CEOs admit not adopting GenAI across their companies in the past 12 months. Why?

1. The unprecedented rate of change with GenAI: it’s hard to board a moving train

GenAI has seen extraordinary adoption — the fastest documented adoption of any new technology in history. One reason may be that because the interface of GenAI is natural language, it’s easily accessible to people. Little education or upskilling is required to use it, unlike with other new technologies, leading to broad, global acceptance. However, it is still consumer-led. Enterprise-grade technology isn’t yet broadly available, which has led to a lull in organisational GenAI adoption.

2. A need for more practical understanding

There’s a gap between knowing GenAI's potential and understanding its specific applications within an organisation. As one client recently put it, “Yes, but what does it actually mean for my business?” This isn't a surprise; it’s a new technology, after all, but it’s a gap that needs to be quickly bridged.

3. Asia Pacific is yet to have uniform availability of technology across the region

Technology is playing catch-up with the pace of the demand. Access to Microsoft Azure OpenAI for example (which provides REST API access to OpenAI's powerful language models) is currently limited due to high demand. There are limits and quotas by country and region. In countries that are able to use OpenAI’s ChatGPT, the data is still only based in the US. However, Microsoft and other hyperscalers are working quickly to improve availability. Once these improvements are in place, we believe there will be the opportunity for Asia Pacific to leapfrog, as it did with 3G to 5G, and gain a competitive advantage.

4. The appeal of GenAI varies, favouring territories with higher productivity and growth opportunities

The CEO Survey revealed the trailblazers in GenAI. Australia leads in GenAI adoption with a rate of 63%, followed by Japan (50%), India and New Zealand (both 39%). GenAI can enhance productivity, streamline processes and contribute to economic growth. This makes it particularly appealing to territories with higher productivity and growth opportunities, with the possibility of GenAI integration into multiple sectors ranging from healthcare and finance to manufacturing and logistics.

5. Risk, compliance and ethical concerns

Half of the CEOs surveyed were concerned about the rising cybersecurity risk (49%), though to a lesser extent than their global counterparts (64%), and 44% expressed concern about GenAI spreading misinformation, compared to global CEOs at 52%. It’s clear that business leaders in Asia Pacific are wary about GenAI, but this also reveals possible gaps in awareness that merit attention. With the launch of new GenAI tools like SORA — OpenAI's text-to-video generator revealed in February 2024 — concerns are likely to increase. Ethical and societal implications commonly listed include fraud, propaganda, and misinformation.  Although many organisations recognise the need for Responsible AI to help combat risks, they’re at different stages of the journey.

Steps you can take to move from intent to adoption

Will the promise of GenAI outstrip reality? It’s clear from our conversations with other technology leaders that many of us consider this to be the first technology that will actually keep up with the pace of the hype.

So, how can business leaders lead from the front foot? Here are several actionable steps you can take:

1. Embed AI horizontally, so that it underpins your whole organisation and other priorities
   
   It can’t just be one vertical. We are often asked how PwC can help introduce a vertical — a new business unit powered by a new piece of technology or software, most likely overseen by a Chief Technology Officer — to meet their needs. Unfortunately, with GenAI, this is like asking, “Where should we use computers?” Everywhere. There are some exceptions, but generally, GenAI technology can help in multiple ways across an organisation. And this is why forming a  strategic partnership is so important — to help work out where it will work best for your business and how to embed it successfully.
   
   The process we’ve found that resonates best with clients follows a five-stage journey: Leadership education and understanding; Strategy development and risk appetite determination; Responsible AI and ethical guidelines; Technology deployment/pilot establishment; and organisational change and upskilling.
   
   
2. Leadership, at the Chief Executive or Chief Operating Officer level, is vital.
   
   The tone needs to be set from the top. The impact of GenAI on an organisation can be profound. It shouldn't be delegated to a Chief Technology Officer.
   
   Then, embed people who are natural early-adopters or ‘activators’, as we call them at PwC, across your organisation. Their job is to help with change management efforts — encouraging its use and sharing the benefits of the new technology.
   
   
3. AI strategy has to be industry and business-led, not technology-led
   
   Not all industries or businesses will see the same value from GenAI. It’s more applicable to some than others. Where, in your specific case, can GenAI bring advantage? Its value should be seen through the lens of your business strategy and how GenAI can help you achieve it.
   
   Your people (closer to the work being done on the ground), as we covered in our recent Hopes and Fears report, can often be a terrific source of ideas on where it can add value. Equally, harnessing their existing excitement around GenAI and creating a culture of experimentation will encourage employees to feel safe and supported enough to step outside the status quo and stimulate innovation.
   
   
4. Look at your planning and funding cycle
   
   The convention of having a three or five-year plan and six-monthly budgets needs a rethink. The pace of change with AI is so rapid that organisations need to be nimble in return. Equally, moving beyond pilot trials to scaling those of value must happen much faster than before, with more immediate Board approval.

How has PwC approached GenAI?

When GenAI first burst onto the scene, as a professional services firm, we had an inkling of its possible power, but its global adoption and speed were startling. But it’s the ability to pivot, adapt and evolve that makes an agile organisation. 

We convened a group at both a network and regional level, with leadership leaning-in, right from the start. We recognised the need to go at this opportunity at a different speed and scale than was customary. We're committed to investing $1 billion globally to expand and scale our AI capabilities, including the services, training and technology we offer clients, to apply AI safely and responsibly, at scale.

It’s this focus that saw us on the stage at the World Economic Forum in Davos last year, to create an assessment tool for clients and develop our own large language model ChatPwC.

ChatPwC was launched firmwide as a pilot in February 2024. It is our safer, secure alternative to public GenAI tools, aligned to our risk appetite and optimised for use in PwC. The platform contains safeguards that help us meet our policy obligations and market and client expectations for applying AI safely and responsibly. It also contains features that extend beyond public-facing AI tools, such as the ability to interact with documents, images and specialised knowledge bases. It's early days, but it will allow us to spend more time on high-value activities that require judgement, creativity and expertise.

Although it’s ChatPwC that has hit the headlines recently, there isn’t any one Generative AI solution or strategy that best meets all of our people's and clients’ needs. We work with several strategic partners - including Microsoft, OpenAI and Harvey - to bring the best enterprise-grade Generative AI platforms and foundation models into the firm, and we augment them with additional PwC-specific capabilities, experience and methodology.

Where next?

The leaps and gains achieved by exercising the full power of GenAI will bring extraordinary change and innovations beyond our imagination today. And those that are early adopters will reap the reward. That makes overcoming inertia on GenAI using early-day AI strategies like the ones we’ve described, even more important.