AI is creating new headaches
Getty
Corporate executives are reportedly becoming worried about artificial intelligence and its potential impact on their businesses.
“Public company executives say the rapid adoption of AI in the workplace is what's keeping them up at night,” wrote Emily Glaser in a recent Wall Street Journal article highlighting executives' growing concerns about AI. “Some point to recent concerns about employees putting their own code into ChatGPT, or companies using generative AI to serve content incorrectly, or worry about so-called hallucinations, where generative AI produces false or inaccurate information.”
Of course, there's no turning back, which means business leaders need to balance the risks and potential benefits of AI. It seems we've been led to expect too much, too soon.
“AI's long-term potential remains significant, but a combination of aggressive marketing by AI vendors and positive public opinion has led to inflated expectations for the emerging technology,” Flavio Villanustre, global chief information security officer at LexisNexis Risk Solutions, told Forbes. “AI still has significant limitations, including environmental impacts due to its high power demands, underperformance due to illusions in many areas, and other issues typical of early-stage technology evolution.”
Given all the hype around AI these days, especially generative AI, it's not surprising to think that AI is a miracle drug for business growth. “Investors are eager for return on investment, and some expect AI to deliver big results for most use cases,” says Diane Gutiu, PhD, vice president of analytics, AI and machine learning at CGI. “The reality is that AI is just another tool — a powerful, but powerful, tool that can solve problems that were previously too complex and costly, like document and image inspection.”
The key is to “use AI at the right time,” Gutiw continues. “If a business problem requires automation, there is a place for AI and it should be considered when considering solutions. By focusing AI on solving strategic problems, organizations will see ROI.”
Business leaders' fears about AI stem from “a mismatch between the benefits companies expect from AI and what the technology can currently deliver,” Villanustre says. On top of that, “AI technologies have limitations that can impede business processes and expose organizations to significant risk.”
For example, increased “regulatory compliance when AI is used to manage personal information” poses risks, he said: “If law requires organizations to erase records of consumers who want privacy, this is not a simple task with current AI models. Erasing a single record would require training an entire model, which could take days or weeks and would be very cost-prohibitive.”
More evidence of return on investment might help ease executive fears, despite all this complexity. “The biggest barrier to AI adoption is proving its business value,” Gutiw says. “Focusing on the problem at hand, rather than the glamorous idea of using AI for every use case, can give companies confidence that the investment will be well worth it.”
A deeper understanding of AI's capabilities and its shortcomings is essential. “Companies need to make an effort to understand AI and identify appropriate AI models before deploying it,” says Villanustre. “Like any tool, AI can have a significant return on investment if used properly, but can also pose risks and losses if used incorrectly.”
Governance is also necessary to maintain alignment with business needs and requirements, measure results, and ensure guardrails. “Organizations achieve some of the greatest benefits by establishing clear AI and data governance to maintain and scale their solutions,” says Gutiw. “Focusing AI on solving specific problems with measurable outcomes makes it easier to measure and monitor its reliability, robustness, and relevance.”
Ultimately, AI's success “should be based on whether a business problem is solved and what benefit AI has brought,” she added. “There should be a clear benefit to solving the problem with AI, such as cost, time or resource savings.”