Generative AI and Professional Services: How AI is Reshaping the Future

Generative AI and Professional Services: How AI is Reshaping the Future

Now that we all know a little more about Artificial Intelligence and generative AI, this blog post is going to take a look at the intersection between those technologies and professional services industries. And to do that, we need to first look at what these two fields offer clients.?

Professional services go back a long time (three of the big four can be traced to the 1840s), so most people are likely somewhat familiar with the industry. But, to give a light outline of the value model of professional services, it offers clients: knowledge, research and experience, asset acceleration (faster time to value), predicting industry trends and comparing cross-industry analysis, benchmarking and business case modeling, problem solving and creativity, relationship management, communication and influence, project management, solution architecture and development, generation of regulatory and compliance strategies, and change management and adoption. They do this in exchange for fees typically paid on a time and materials, fixed-price/fixed-scope, or sometimes an outcome/risk-based, commercial model.?

But what about the capability model of AI when applied to professional services? I’ve ignored a number of capabilities that I personally believe do not mesh well with professional services.? I’ve organized the list into three layers based on my perception of the degree of ‘human support’ (review or intervention) required.

First we have the basic tier, the least likely to require human support, which includes: analyzing, comparing, classifying, monitoring, pattern matching and recognition, scaling and segmenting. Then we have the intermediate capabilities, that may require human support, that include: answering, augmenting, automating, predicting, recommending, relating and clustering, and translating. And finally, we have the advanced capabilities, those most likely to require human support, including: arbitrating, attesting or verifying and evaluating, generating, synthesizing, and visioning and composing. With the capabilities of both professional services practitioners and AI, the real question is of how they work together to support better client outcomes.??

Those reading my descriptions of the professional services and AI models closely may have noticed some overlaps between the capabilities —?and this intersection is where the disruptive change occurs that the industry should prepare for.

To put this capabilities intersection into context, I’ll use the legal profession as an example. Full disclaimer —?I’m not a lawyer. I did, however, study law when I was qualifying as a chartered accountant, and the one thing I remember from this time (other than my outstanding teacher) is the need to learn, recall, and apply case precedents to points of dispute. This aspect of the law can be one of the hardest, with the overwhelming volume of data coupled with limited human ability to recall, recognise and match these precedents to points of view. But, taking a look at the model above, these tasks fall into the “Basic” AI capabilities.?

So, how can these AI capabilities be applied to drive better outcomes for clients? A few basic examples (not limited to the legal industry) include using AI to schedule appointments, or creating chatbots to route queries to the right experts and even provide basic guidance on simple questions. But, the ability of AI in the legal world could go even further. Could these ‘digital paralegals’ perform risk analysis and predictions, while generative AI could create initial position papers or contracts. Meaning, many processes could be automated and accelerated, allowing a legal professional more focus on creating a case for their client. Then there’s the significant productivity gains. A study overseen by Stanford Law School indicated an average time of 92 minutes for a human lawyer to review 5 Non-Disclosure Agreements versus 26 seconds for the AI lawyer. The most shocking part? That study is now 5 years old.???

But, the legal professional is not typically known as an early adopter — leading to a risk of missing the potential power of the technology (especially given the velocity at which AI is evolving). Imagine, instead, a reality in which courtroom lawyers had an AI-based tool listening in real time to a trial or cross-examination, instantly suggesting potential precedents, questions, or advice based on a particular judge’s historical responses to legal arguments or tactics. Not even Della Street would be able to compete with that game-changing AI assistance.??

These benefits also bring potential disruption to existing business models. If tasks normally done in hours or even days now, in an AI-assisted world, take minutes, the legal profession faces a challenge in monetising this change. Billing in 6 minute increments might no longer apply. Plus, fee structures are often predicated on a highly leveraged rising scale of expertise (going lowest to highest) from the legal clerks, to paralegals, associates, and senior associates through to partners and managing partners. Automating might flatten the structure. So, what happens, then, to a firm’s fee base?

There’s also a potential risk of client attrition as use of generative AI tools like ChatGPT as lower-cost pseudo-lawyers becomes more widespread. While the risk of hallucination (i.e.the generative AI actually is not accurate) is significant, there are already simple examples where ChatGPT has been used to mitigate, or litigate, parking tickets and minor cases like landlord disputes. Small Claims Court could simply becomes the submission of each party's relative position, which AI then adjudicates with generative settlements, outcomes and advice —?this may be less entertaining than Judge Judy, but it’s a lot more efficient.?

Arguably, humans will always be stronger at the emotional side of practicing law, as it is a social practice. Though it remains to be seen (and can be a challenge for the “next frontier”) if an impassioned plea to a jury of peers can ever land quite the same when a machine delivers it. But, as AI safely takes on routine tasks, the role of lawyers and legal professionals can shift even more towards complex problem-solving, negotiation and relationship-building with clients. For higher level capabilities, I anticipate humans will need to act as experts and reviewers of AI generated outputs to counter the risk of machine hallucination. And while some in the legal and professional services industries might be wondering, “Where will the next generation of experts come from if not trudging their way through years of ‘grunt’ work? How will the next generation of experts actually get their expertise? What will the basis of their experience be?” —? my dear friend and former client, Bob Richardson, always told me, “Mark, experience is what you get when you don’t get what you expect.” He was right.

Many of the ideas and impacts discussed in this paper apply outside the legal profession. In fact, any professional services organization predicated on a leverage skills/experience delivery model and particularly those using time and materials-based billing, or those with a significant portion of their offerings found in the “Basic” and “Intermediate” AI capabilities above. Now is probably a great time to begin thinking about how to embrace and integrate this technology to serve clients more effectively.?

In the next post, I’m collaborating with my friend and colleague Aaron Tunesi, who leads our Retail, Consumer Goods, Travel, Transport & Hospitality business in the UK for Salesforce and Chris Poole from our UKI Maverick Team, to uncover the impact of AI on project management. Until then, and as always, your thoughts, input and feedback in the comments below are welcome.

Mark Wakelin

Disclaimer: The views expressed in this article are mine alone and are not those of Salesforce