How ChatGPT and generative AI will change the digital workplace

April 5, 2023 Updated: March 6, 2024 by

Once every few years a technology emerges that will change the digital workplace. The invention of the iPhone transformed the way all of us consume information. The emergence of Yammer helped to accelerate the spread of social technologies in the enterprise. And right now, the release of ChatGPT feels like one of those step changes that is going to influence the landscape of work technologies in a profound and potentially rapid way.

Of course, artificial intelligence (AI) and machine learning (ML) are not new and have been bubbling along steadily for the past few years. While the debate about AI’s role has sometimes strayed into utopian and dystopian themes, in the background, models, approaches and ultimately products have been rapidly advancing, including in the areas of generative AI.

The rise of generative AI

McKinsey defines generative AI  as “algorithms (such as ChatGPT) that can be used to create new content, including audio, code, images, text, simulations and videos.” There is already a market of products that are already helping to produce text, images, videos, music, voices; and these are already being used by organizations and teams in their daily work.

Meanwhile there has been significant progress in the development of large language models like ChatGPT that can interpret and generate text.   

ChatGPT is therefore not necessarily new, but its ability to generate sophisticated text in response to natural language queries and instructions has captured our collective imagination in a way that no other AI-powered tool has yet done so far. Many people experimenting with ChatGPT have been very impressed when they first experience using it – “I’m blown away” is a common response. It’s opened up our collective imagination of what is possible with generative AI – and at the same time people are actively using it in and outside work.

The potential of ChatGPT is huge…

The potential to use ChatGPT to generate intelligent text within the workplace to increase productivity is tantalising. From producing minutes of a meeting from an automatic transcript to generating a non-disclosure agreement (NDA) to order to simply tidying up a page of content – all within seconds –  there are hundreds of potential uses in terms of producing output.

But then the potential goes much further with ChatGPT acting as a search and reference tool, automating content production at scale, supporting software development by producing or suggesting code and bringing new layers of sophistication to robot process automation. And this is just the start, because as ChatGPT and other large language models become more sophisticated and combine with other services, the opportunity for innovation could go much further.

The potential has captured the imagination of businesses, and some are already putting ChatGPT to use. For example, a recent report from Accenture based on a survey of 5,000 global executives found that 96% were “very or extremely inspired by the new capabilities offered by AI foundation models.” The report also cites an example of CarMax, a vehicle seller, using ChatGPT3 to analyse 100,000 customer reviews and auto-generating 5,000 readable summaries, an activity that would have taken their existing human content team 11 years.

..but here are the inevitable risks

For every exciting possibility that ChatGPT comes with to revolutionize working practices and turbo-charge productivity, it comes with a range of enormous potential risks. These risks are heightened because we are also still in a period of collective learning, the platform is still evolving and approaches are still unclear. The hype around ChatGPT and other generative AI has also tended to paper over some of the cracks that appear when you look more closely at the output from ChatGPT and find problems with accuracy.

One of the key risks is around data privacy and GDPR. Any private information submitted into ChatGPT might also be used to then further train the model and almost certainly does not have an individual’s consent. If, for example, I have some text that I want ChatGPT to clean-up, it may include personally identifiable information. This is a very grey area and thoughts and practices are still emerging. Italy, for example, has temporarily banned the use of ChatGPT largely down to data privacy concerns.

The protection of intellectual property rights is also another area where the lawyers are waiting in the wings. ChatGPT has been trained by publicly available sources, but it doesn’t provide any information on what those sources are. An answer provided by ChatGPT could be based on copyrighted material. Here the dilemma for organizations is firstly, how do I protect my own intellectual property, and secondly, who is responsible if copyright has been breached?

A further key concern is that information on ChatGPT is not up to date and it is not necessarily correct. There are examples of “hallucination” where it appears ChatGPT has dreamt answers up as they are completely inaccurate. A local mayor in Australia is already considering action against ChatGPT for defamation  for inaccurately stating he was imprisoned for bribery. This leaves questions about not only relying on information from ChatGPT for decision-making but also questions over liability of producing output based on it. One of the problems is that ChatGPT presents information in an authoritative and confident way, and so it is becomes very easy to believe something is a fact when its actually not.

There are also many questions of what using ChatGPT does to some key internal processes. If I use ChatGPT to generate an NDA for example, does that present a risk to my firm, rather than using an approved example? If I generate a blog post using ChatGPT does that actually start to undermine the power of personally written content and make it more bland and generic? And if rely on ChatGPT too much does that mean I fail to learn about key processes if I am a new starter? There are also wider risk and ethical questions about the use of ChatGPT, for example with the inherent bias within its output.

And finally there are also wider concerns about the consequences – some unintended of using ChatGPT. Will it lead to a loss of jobs? Can it be used to undermine cybersecurity? Even does it upset and disrupt our fundamental business model? And where does it lead to in the future? Here there are genuine fears, even leading to an infamous letter signed by Elon Musk and other influential technology leaders suggesting a six-month pause on AI research, although the letter has since been revealed to have some flaws.

Trying to get the balance

In a time when ChatGPT and generative AI is moving so rapidly, business leaders, IT functions and digital workplace teams are trying to work out how they can balance the existing possibilities of using ChatGPT while also reducing the risks. The need to provide some clarity does have a sense of urgency because many employees are actively using it in their everyday work and there is a sense of trying to gain competitive advantage. At the same time, the legal and regulatory approach to ChatGPT is unclear and being actively considered, and could bring the use of generative AI to a sudden halt.

The kind of questions that internally teams are asking include:

  • What processes can we improve using ChatGPT?
  • What can we action now?
  • How can we use it in the short-, medium- and long-term?
  • How can we harness ChatGPT for innovation?
  • How can we reduce the risks?
  • Should we ban or encourage the use of ChatGPT?
  • What governance and related guidance do we need to put in place?
  • What are the broader ethical consequences?

Of course, the responses from organizations have been different. Some have decided to ban the use of ChatGPT outright, others have produced or are working on detailed guidance, and some are actively conducting experiments. And in large complex organizations sometimes all three of these approaches are likely to be taking place at the same time, albeit in different parts of the organization.

Eight ways generative AI will change the digital workplace

Inevitably some of the short-term challenges to using ChatGPT and other powerful generative AI services will be ironed out, even if navigating the opportunities and risks remains an ongoing challenge. In the medium to longer term, generative AI is likely to change the digital workplace in a number of different ways. Below are eight areas where we think this will happen, but there could be additional areas, and potentially change may not be as profound as we think.

1. Generating content

Generative AI has the potential to change the processes for generating the content that we use across the digital workplace. Arguably, it’s going to make it much easier, reduce the time required, lower the barriers to content production and help optimize content for elements such as findability and readability.

It’s likely that those organisations that can most effectively integrate the use of generative AI in the content production process without stifling or reducing the creative and expert human input required to make that content valuable and purposeful, will be the most successful.  

2. New digital skills and roles

ChatGPT and other generative AI present significant opportunities to increase productivity, but to get the best out of them requires new digital skills. For example, for end users this might be in how to:

  • construct queries and requests to get the best outcome
  • recognise information that is incorrect
  • navigate multiple areas of risk.

There are also likely specialist roles that can benefit from using generative AI, as well as new roles that will be created. Organizations will want to also build their own custom applications for internal and external use. Those organizations that can build up the specialist knowledge around the use of generative AI while also enabling wider digital literacy will be in the best position to leverage the use of ChatGPT and similar services. Digital workplace teams can be hugely influential in this respect.   

3. Governance

There is an urgent and obvious need for governance in the use of generative AI that balances the advantages of using tools like ChatGPT to drive productivity while also reducing the considerable set of risks already around data privacy, intellectual property protection, relying on unreliable data and more.  Digital workplace teams have a key role to play in establishing and implementing these policies, bringing a balanced and pragmatic view to the table, where different stakeholders will have differing views. As it is likely that governance measures will rely on user trust, there is a strong link here to the digital skills and literacy area.

4. Digital workplace applications

Inevitably ChatGPT and similar services are going to be increasingly threaded into different digital workplace applications, especially as more vendors hook into the ChatGPT API or some of its competitors. With Microsoft’s huge investment in OpenAI, this will include Microsoft applications with the launch of Microsoft Copilot, with the power of generative AI weaved right into Word, Excel, Outlook, Microsoft Teams and more. This again means digital workplace teams have a significant job on their hand to support the roll-out, governance and best use of these applications, with likely more rapid change on the horizon.  

5. Low-code no-code solutions

ChatGPT is also likely to particularly advance low-code no-code solutions where non-IT professionals are able to create simple apps, workflow and automation without the help of the IT function. For example, ChatGPT is already producing code and there are examples of people starting to design simple video games. The extent of what non-IT professionals can achieve may well expand, again which may come with a new swathe of governance challenges.

6. Findability

ChatGPT has multiple possibilities to transform findability within the digital workplace. A sophisticated chatbot like ChatGPT may well prove to be the interface with which people try to find items, but it may also be that the emphasis on findability moves to finding information and answers buried within sources, rather than the original sources themselves.

The generative aspect of ChatGPT also could be used to make existing content more findable, for example by doing the heavily lifting by automating tagging, generating smart summaries and more which helps bring the content people need.

Outside the world of enterprise search, Bing is being injected with ChatGPT, while Google is still working on implementing Bard, it’s answer to ChatGPT. The evolution of popular search engines has an enormous influence of employee expectations about findability in the workplace as well as the product roadmap of enterprise search applications, so this is another area where AI is likely to influence the direction of findability across the digital workplace.

7. Expectations

ChatGPT has alerted many to the potential of AI. It seems likely that it will increase the expectations of business stakeholders and employees in what they think their digital workplace and the related digital employee experience should be. This means it could raise the bar on what users want to get out of individual products and the kind of interfaces they want to interact with. Unfortunately, there’s a likely disconnect between expectations and the reality of what can be deployed, something we already see in the world of enterprise search.

8. Change will the exponential

The pace of change in the digital workplace is showing no signs of slowing. With the evolution of ChatGPT and other AI services, the pace of change is now likely to be exponential as new services are defined, innovations discovered and other services pushed to integrate generative AI. The combination of other emerging technologies like the Metaverse with generative AI also has enormous potential. What we can do tomorrow is potentially mindboggling – with advances we can’t predict. Digital workplace teams already struggle to keep up with implementing new products and services; that job is not likely to become any easier.

Digital workplace teams can play a role

The potential for ChatGPT is huge, but so are the risks. The digital workplace is likely to evolve in ways that we don’t even know yet, and that pace of change could be rapid. Digital workplace teams have an important role to play in all this, navigating the balance between innovation and risk, use and governance, and the hype and reality. I’ve got a feeling it could be a rollercoaster ride.

Talk to us about… 

… becoming a DWG member

More about DWG Membership »

Categorised in: Digital workplace trends, Future of work

Steve Bynghall

Steve Bynghall is a freelance consultant, researcher and writer specializing in the digital workplace, intranets, knowledge management, collaboration and other digital themes. He is DWG’s Research and Knowledge Lead, a benchmark evaluator and research analyst for DWG.

Don't journey alone

Why choose DWG?

Learn more about DWG and our history, and the benefits of working with us.

Read More
Your digital journey starts here

Book a free one-to-one consultation to discuss the current status of your digital workplace. Each consultation is followed up with a bundle of useful resources to help get you started.

Book your consultation