Ignoring Responsible AI Could Cost Orgs at Least $1M
Ensuring that AI vendors, practitioners and researchers follow ethical AI practices is about more than social responsibility; it's also about money.
That's according to more than a third of respondents in a recent survey commissioned by Amazon Web Services (AWS). The cloud giant, working with research firm Morning Consult, polled U.S.-based business leaders on their respective companies' views and plans around responsible AI, which it defined as "designing, developing, and using AI technology in a way that is fair, accurate, secure, transparent, safe, and inclusive."
Though the companies differed on when (and to what degree) they plan to get serious about responsible AI, a good chunk of them said not doing it at all would put them at considerable financial risk. Specifically, 35 percent estimated "it could cost their company at least $1 million -- or potentially jeopardizing the business itself ."
Just 17 percent of respondents said ignoring responsible AI wouldn't cost them a cent.
"Many businesses (75%) see the benefits of deploying AI, especially in specific areas such as increasing revenue," AWS said in the report, but added, "[B]usiness leaders know they have to do it right."
As for when they'll actually do responsible AI, most organizations seem to be dithering. While more than three-quarters (77 percent) of survey respondents professed knowledge of responsible AI and more than half (59 percent) consider adopting it a "business imperative," only a quarter of them have started to make plans for doing so.
"Further," said the report, "63% of respondents report that their company does not have a team dedicated to responsible AI."
AWS attributed this hesitation to "a number of factors, such as how quickly the technology is evolving and the lack of education on responsible AI."
Speaking of which, businesses are also undecided about whether to provide employee training that's specifically focused on responsible AI; 39 percent said they would and 39 percent said they wouldn't.
Regardless of where they are in the process of adopting responsible AI, a lot of respondents (47 percent) expect it will take up more of their budgets in 2024 than it did in 2023. Only 10 percent said they actually plan to reduce their spending around responsible AI in the new year.
Other interesting takeaways from the survey include:
- There's no consensus about who is actually responsible for fostering responsible AI; 29 percent put the onus on vendors, 27 percent on businesses end users, and 16 percent on academics and researchers.
- Perceptions and knowledge of responsible AI tend to fall along generational lines. Business leaders aged 44 and younger were more likely to report activities around responsible AI (whether that's training, investment, planning or adoption) than those aged 45 and older.
- "AI-powered solutions" are expected to be in use at the vast majority of businesses (92 percent) by 2028.
AWS released the results of the Morning Consult survey to coincide with its recent re:Invent conference this fall. More information on the survey is available here.