Navigating Generative AI: Enterprise Challenges, Strategies, and Adoption Stages

SHORT

In the relatively short time since enterprises began adopting generative AI, they have already started adapting their approach to leveraging this technology. They are now confronting several challenges that have emerged, including the high costs associated with large-scale use of large language models (LLMs), inadequacies in their cloud and data infrastructure, concerns about the reliability of generative AI outputs, and the complex risks related to regulation, intellectual property, and ethics. Rather than relying on a single vendor, enterprises are increasingly using multiple model providers to develop their generative AI applications. Additionally, they are turning to more cost-effective open-source models that they can control and refine, rather than investing in the costly and complex process of building custom models from scratch. This approach aligns with what McKinsey research refers to as adopting the role of a generative AI model “taker” or “shaper” rather than a “maker.”

Despite these widespread changes, enterprises are at varying stages of generative AI adoption. Our analysis categorizes organizations into three stages of readiness and adoption:

Observers: The majority (50-60%) are “observers,” focusing on AI readiness by establishing their data and cloud foundations and conducting small-scale proofs of concept (POCs) for internal use cases, such as text summarization and knowledge management, aimed at enhancing productivity.

Front-runners: Another 30-40% are “front-runners” with a clear vision for using AI and generative AI to scale cost reductions through applications like AI-enabled service desks and legacy-code upgrades. These organizations have restructured their operations by creating AI centers of excellence (COEs) and new roles such as chief AI officer.

Innovators: Less than 10% are “innovators” who are AI-first and have integrated strategies to achieve cost reductions and unlock the full potential of generative AI. They focus on advanced use cases like AI-enabled product design and development to drive future business growth.

In brief enterprises are adapting to generative AI by using multiple model providers and cost-effective open-source models to overcome high costs, infrastructure challenges, and regulatory risks. They are at varying stages of adoption, with the majority focusing on AI readiness, some scaling cost-reduction efforts, and a few integrating advanced AI strategies for future growth.

@McKinsey #GenAI #AI

Artykuł Navigating Generative AI: Enterprise Challenges, Strategies, and Adoption Stages pochodzi z serwisu Norbert Biedrzycki.