Enterprises lose millions each year due to poor AI models

Underperforming AI models, which are built using inaccurate or low-quality data, are resulting in misinformed business decisions.

A study for Fivetran, conducted by Vanson Bourne, polled 550 from organizations with 500 or more employees and finds that on average they lose six percent of their global annual revenues, or $406 million, to poor AI models.

Almost 90 percent of organizations are using AI/ML methodologies to build models for autonomous decision-making, and 97 percent are investing in generative AI in the next one to two years. At the same time, organizations point to challenges of data inaccuracies and hallucinations, and concerns around data governance and security. US organizations leveraging large language models (LLMs) report data inaccuracies and hallucinations 50 percent of the time.

"The rapid uptake of generative AI reflects widespread optimism and confidence within organizations, but under the surface, basic data issues are still prevalent, which are holding organizations back from realizing their full potential," says Taylor Brown, co-founder and COO at Fivetran. "Organizations need to strengthen their data integration and governance foundations to create more reliable AI outputs and mitigate financial risk."

The study finds 24 percent of organizations reporting that they have reached an advanced stage of AI adoption, where they utilize AI to its full advantage with little to no human intervention. Technical executives -- who build and operate AI models -- are less convinced, however, with only 22 percent describing their AI maturity it as advanced, compared to 33 percent of non-technical workers. When it comes to generative AI, non-technical workers’ high level of confidence is coupled with more trust, too, with 63 percent fully trusting it, compared to only 42 percent of technical executives.

Data is still a stumbling block too, with most organizations struggling to access all the data need to run AI programs (69 percent) and cleanse the data into a usable format (68 percent). New generative AI use cases have introduced further complications, with 42 percent of respondents experiencing data hallucinations.

The full report is available from the Fivetran site.

Image credit: aremafoto/depositphotos.com