Balancing Opportunities and Risks
Generative AI holds tremendous promise for businesses, but also comes with significant risks that are increasingly apparent. Machines capable of performing an ever-larger share of (white collar) tasks promises first and foremost significant efficiency gains across the value chain of many industries. Here, it picks up where lower-level forms of machine learning using models like Decision Trees or Support Vector Machines have left off. Second, generative AI also offers opportunities for companies to differentiate themselves from competitors in their offering or features. For example, the integration of AI into Bing’s search console, which can answer more complex search queries of users than before, enabling Bing to expand its search market share historically dwarfed by Google's. Third, delegating mundane tasks to machines can free up capacity for higher-level tasks, transforming professions at a high pace.
However, generative AI models have also sparked controversy for several reasons. One major concern is the lack of transparency in the training data used for the models, which raises the risk of infringing on intellectual property rights. Additionally, unguided use of the technology by employees can lead to leaks of a company's own intellectual property or sensitive information, as demonstrated by the recent code incident at Samsung. Another concern is the inability to trace the input parameters or training sources that led to a certain outcome. This lack of transparency raises the risk of generating fictitious answers (“hallucination”), as the model runs a probabilistic algorithm to generate output. Furthermore, the rapid acceleration and wider adoption of large AI models may have disruptive potential on society, as parts of certain professions are affected by automation. If history serves as any predictor, this fear is as old as technical innovation itself and every wave of automation has led to the creation of new jobs while lifting productivity and prosperity.
While further understanding and maturing of the tools (and its guardrails) is likely to address some of those fears, governance through regulatory bodies will have an important role to play.
Emerging Rules of the Game
While generative AI technology itself is not new, its accelerated development has put regulators in a challenging position to provide a level playing field that safeguards the interests of all stakeholders. Historically, the US and UK have taken a more liberal approach, incentivizing innovation, while Europe has prioritized consumer protection and competitive markets.
Regardless of the approach regulators choose, it's fair to assume that the rapid development of generative AI will continue, and market participants must take this into account when making investment decisions and strategic choices. We can draw analogies from the shared economy (with Uber and AirBnB capturing dominant market positions before disputes with incumbents and municipalities were settled), autonomous driving and accounting for emissions cost, which show that early adopters can gain a competitive advantage.
Implications for Transaction Advisory and Value Creation
Against the backdrop of rapid development in generative AI, investors and executives are increasingly concerned with two questions:
- What is the impact of this technology on our company or portfolio?
- How can we put (generative) AI to work?
For an impact assessment, both risks and opportunities for a business should be considered. From a risk perspective, it is important to consider the potential disruption of the business model. To assess this risk, a systematic analysis of the value chain and key processes is necessary to identify which parts can currently and in the future be performed by AI. It is also important to examine how the deployment of AI may affect the company's suppliers or customers, as disintermediation may be a concern in some cases.
On the other hand, from an opportunity perspective, we would ask how the deployment of tools can decrease input costs, automate manual tasks, or enable new products or features that provide a competitive edge. Especially, companies’ proprietary data sets, which can be used to train own models, can be supercharged by AI to an even stronger source of competitive advantage.
The deployment of AI in an organization requires an initial strategic analysis of application fields along use cases (“where to play”) and an assessment of the organization’s AI readiness (“how to win”). The assessment of use cases can be structured along the key activities of the business along the value chain and then prioritized by the strategic impact (e.g. cost savings potential) and the ease of implementation (e.g. determined by the availability and maturity of existing solutions). Additionally, the AI readiness assessment should investigate data capabilities, tech infrastructure and processes, and partner ecosystem maturity.
To then bring AI to work in business reality, an agile development approach should be used to build an MVP and enable fast iterations and learning cycles. To operate the AI tools effectively and responsibly, a modified operation model should be established, potentially including modified processes, roles and responsibilities, and governance to safeguard the company's and stakeholders' interests.
Do you want to know more about our expertise? Get in touch!
Development and Execution of a Customized Digital Growth Strategy