Businesses across the globe are frequently ignorant of the threats generative AI (GenAI) poses to cybersecurity and data privacy, new research has suggested.
The study by US cybersecurity company ExtraHop surveyed 1,200 IT and cybersecurity professionals from around the world and found that inaccurate responses, or hallucinations, were the top concern for businesses using GenAI tools.
Around 40% of respondents answered that this was their main concern, whilst only 36% of respondents cited security-centric issues as their top worry.
These security concerns included the exposure of employee and customer identifiable data.
Despite a large swathe of companies having banned GenAI tools such as ChatGPT, ExtraHop’s study concluded that many business prohibitions have proven ineffective.
Whilst a third of businesses who took part in the study stated that they banned GenAI tools, only five percent of respondents stated that their employees never use these tools at work.
In an industry update on managing the dangers of AI, GlobalData principal analyst Steven Schuchart wrote that whilst AI cannot be undone or “stuffed back into the bottle”, its rapid growth remains largely positive for businesses so long as they maintain good cyber hygiene and ethical practices.
“There is a serious need for robust laws about how AI can be used,” Schuchart wrote, “This must include comprehensive strengthening of individual privacy and personal data sovereignty.”
Alongside strict and clear legal guidance, Schuchart also called for AI to be auditable and transparent, calling for an end of “black box” AI development that shrouds training data in mystery.
Despite Schuchart’s warnings over correct AI uses, ExtraHop’s survey found that only around 46% of businesses had AI governance policies in place, and a further 42% trained employees on how to use AI tools.
Despite this, GlobalData’s own research suggests that 17% of businesses have already integrated AI into their workdays citing a high level of adoption.
Training employees on proper AI use and risk management of generative AI should be a top priority for businesses eager to implement the technology.
