Talk about the endless possibilities and impact of artificial intelligence is rampant, yet a new report by Everest Group shows that the majority of enterprises (83 per cent) are currently only testing the capabilities of AI through pilot programs, or have adopted generative AI for one or more production-grade use cases.
“Our research clearly documents that most organizations are in what we call ‘Wave 1’ or the pilot phase of Gen AI adoption; however, in 2024 and 2025 we fully expect more organizations to advance to the ‘Wave 2’ phase of production-grade deployments,” said Abhishek Singh, partner at Everest Group.
Wave 2 of generative AI adoption, the report says, is when we’ll see enterprises move from small-scale pilot projects and experiments to enterprise-wide scaled pilots, and increased focus on optimizing performance of generative AI models, as well as the initial adoption of enterprise AI platforms.
As part of this study, more than 50 chief information officers were jointly interviewed for their perspectives on current adoption maturity and key strategies and challenges, as well as future investment plans in Gen AI.
Over 60 per cent of feel that the fast-evolving and confusing technology landscape is one of the top challenges for them when scaling their generative AI initiatives.
Canadian C-suite members echo the same sentiment. A recent Global Leadership Monitor survey by Russell Reynolds Associates (RRA) showed that one in three leaders in Canada say that they are uncomfortable with implementing generative AI.
Globally, 55 per cent of leaders say that knowledge and expertise are top barriers to implementing generative AI, and 72 per cent agree that a strong understanding of generative AI will be required for future C-suite members, the same report affirmed.
CIOs also cite things like a lack of clarity on success metrics, budget constraints, talent shortage, and data security and privacy concerns as top barriers to scaling AI, the Everest study indicated.
As we look ahead to 2024, leaders need to get on board, RRA said. They need to come to terms with the fact that generative AI is here to stay, and will impact the way we do business and which companies will stay ahead of the game.
The three generative AI areas that have gained substantial adoption, according to Everest, are:
- Content creation and preparation – creative writing, email generation, language translation, etc.
- Knowledge management – context-aware search, summarization, conversational employee interface
- Software development – automatically generating code snippets, scripts, or even entire programs based on natural language descriptions or high-level requirements
All these use cases have converged and taken over areas such as customer service delivery and management. A report by Zendesk reveals that 70 percent of customer experience (CX) leaders are reimagining their customer journeys using tools like generative AI, and a staggering 83 per cent report positive ROI.
Everest also highlighted other generative AI use cases that are in the exploration stage. Financial institutions, for instance, are exploring things like financial bots, use of synthetic data for risk simulation, claims processing and more. The healthcare industry is looking to use AI for medical report generation and drug research and discovery, while the media and entertainment industry is testing AI for game development, AI avatars, as well as AI-generated media posts.
The “Wave 3” of AI adoption, slated for 2026 and beyond, the report claimed, will see enterprises innovate and create custom-built generative AI solutions to meet specific business needs.
Navigating this journey, CIOs will have to pay attention to the following four key considerations, Everest affirmed:
- Generative AI is an expensive technology, so it is imperative that leaders establish and translate the right business objectives into the right value equation to create a successful Gen AI strategy.
- Assess their digital maturity to lay out the right AI adoption strategy – They need to firstly prime their data foundation, making sure it’s fine-tuned to the context of the enterprise. Additionally, enterprises need to ensure their talent is AI-ready. Employees need to understand the technology, its implications and applications.
- The countless risks of generative AI, from cyber threats, data privacy violations, to AI hallucinations, biases, and risks to intellectual property require enterprises to implement a well-structured, enterprise-wide risk and governance strategy. They need to ensure, for instance, that there are guardrails for handling and storing sensitive data used in training and inference, there is transparency embedded in their AI systems, that they stay abreast of the evolving regulatory environment and conduct regular compliance audits.
- Enterprise leaders need to select the best foundation model vendor for their generative AI strategies. That entails careful assessment of the vendor’s expertise, particularly evaluating the model’s performance on relevant benchmarks, adaptability, data security, and privacy policies, as well as ensuring alignment with regulatory requirements.
The full Everest report is available for purchase here.
