A survey of 1,003 executives in the U.S, that work for organizations with more than $100 million in revenue, published today, finds nearly all (97%) plan to build or buy generative artificial intelligence (AI) solutions to automate business processes or create new revenue streams, with 95% reporting they trust publicly available tools and services.
Conducted by Atomik Research Insights & Analytics on behalf of Fortanix, a provider of a data security platform, and Intel, the survey also finds 88% work for allocations that have assigned budget for generative AI technologies, with nearly three quarters (74%) reporting they are under pressure to deploy it imminently.
More than a quarter (28%) said they plan to build their own platforms and services compared to 36% that will prefer to buy. Another third (33%) expect to rely on a combination of building and buying.
Overall, the greatest benefits organizations envision as a result of GenAI adoption are improved productivity (47%) and cost savings (39%), the survey finds
At the same time, 97% of respondents also noted their organization blocks or restricts generative AI usage, with 89% reporting the controls applied are moderately to very successful.
Nevertheless, the survey finds a third of respondents (66%) already use generative AI for work, with 69% using a work account and 64% using personal email accounts to access these tools.
That dichotomy suggests there is still a significant gap between the level of enthusiasm for generative AI and the potential risks, says Anuj Jaiswal, chief product officer at Fortanix. Many senior business leaders, especially, are not fully cognizant of how, for example, sensitive data involving intellectual property might be compromised, he adds.
In fact, the survey finds that among the 301 security executives that participated in the survey, a full 87% of security executives admitted there had already been a breach in the past 12 months.
As more data is exposed to generative AI platforms and services, the probability there will be a breach only increases, notes Jaiswal.
In fact, those breaches will one day soon highlight how often cavalier organizations today are about the way data is managed across multiple applications and platforms, he adds. “There’s already a lot of data sprawl,” Jaiswal says.
It’s not clear to what degree generative AI might encourage organizations to revisit how data is managed and secured, but it’s only a matter of time before GenAI platforms generate outputs that include data that many business and IT leaders will be surprised has been discovered. Providers of generative AI tools and platforms generally assure end users that if they opt out, their data will not be used to train AI models, while others make no such assurances.
Unfortunately, not every end user carefully reads the end user agreement provided to them. Additionally, there is no guarantee that existing policies might not change.
None of those concerns are likely to slow down adoption of generative AI, but, at the very least, business and IT leaders should come to terms with the fact that for all the promised benefits, there will also be inevitable consequences that are still largely being unforeseen.
