China leads the world in business use of generative AI, with 83% of organizations saying they are actively adopting it. However, using technology and leveraging it to achieve a goal are two different things. And when it comes to full adoption, the U.S. leads all countries at 24% (China at 19%).
Researchers from SAS and Coleman Parkes Research surveyed 1,600 decision makers around the world responsible for GenAI or data analytics strategies and published the findings.
Respondents represented a variety of sectors, including healthcare, and the organizations surveyed ranged in size from over 500 employees to over 10,000.
SAS released its findings and analysis on July 9 and offers tips for effectively incorporating this technology into existing or evolving operations. Here are six:
1. Use data management tools to ensure that your Large Scale Language Models (LLMs) are fed the highest quality data and prompts – data that is auditable and traceable.
These tools provide user privacy and security with strong data protection measures such as data minimization, anonymization, and encryption to ensure sensitive information is protected, the report authors noted. “Furthermore, workflows can be automated to make building and reconciling LLMs the quickest and most direct way.” Learn more:
“Organizations should refer to their governance and compliance policies for key frameworks within which data management tools can be applied.”
2. Before developing a comprehensive GenAI strategy, ensure that key decision makers are familiar with AI.
“This takes time and often requires hiring outside experts to advise the team,” the authors write, with Brian Harris, executive vice president and chief technology officer at SAS, adding:
“With any new technology, organizations need to go through a discovery phase, separate the hype from the reality and understand the complexities of real-world implementation in the enterprise. We've reached this moment with generative AI.”
3. Identify optimal GenAI use cases to achieve rapid return on investment.
The first step to successfully adopting GenAI is to identify use cases that will deliver high impact for the technology, so that a measurable return on investment can be realized as quickly as possible, the authors suggest. Marinela Profi, Strategic AI Advisor at SAS, explains:
“LLM alone doesn't solve business problems. GenAI is just a capability that can extend existing processes, but you need the tools to enable integration, governance and orchestration. And most importantly, you need the people who can use the tools to ensure the right level of orchestration.”
4. Ensure your GenAI software vendor can integrate with your existing workflow and decision-making platforms.
GenAI is an ideal contributor to hyperautomation, facilitating the automation of all tasks that can be performed within an organization, the authors say, adding:
“GenAI excels at summarizing vast amounts of data to support decision-making workflows, enabling real-time interactions aligned with desired business processes.”
5. Incorporate GenAI into existing business processes using decision-making workflow systems to drive measurable outcomes.
The authors point out that LLM can only perform some of the tasks in the use case.
“Organizations still need an end-to-end process to align the AI lifecycle with greater transparency and governance for LLMs.”
6. Prepare for problems.
“For any organization, the use of GenAI can raise concerns about data privacy, security, and lack of governance, as well as concerns about reliance on the technology and its potential to amplify bias,” the authors write. “Many of these organizations are ill-prepared for regulatory compliance and lack the means to oversee GenAI governance or the technology.” Read more:
“Our research shows that companies are rushing to adopt GenAI before establishing proper governance systems, which could lead to serious quality and compliance issues down the line.”
SAS' full research report (contact information required for access) and interactive data dashboard provide much more information.
