To AI Hell and Back: Finding Salvation Through Empathy

A Forrester analyst takes her audience on a journey to the fiery depths of GenAI’s security risks and says technology paradise can be found on the other side.

Dante’s “Inferno,” the epic poem from “The Divine Comedy” that drags its protagonist though nine circles of Hell, proved to be an apt analogy for AI’s sudden explosion of use in the tech industry and beyond as hellish security risks and compliance hurdles await.

But so do heavenly rewards, according to Forrester Principal Analyst Enza Iannopollo, whose keynote closed out the firm’s Security & Risk Event this week in Washington, D.C. It was a fitting conclusion to an event centered on the emerging technology of generative artificial intelligence (GenAI) and its profound risks and rewards.

The public release of OpenAI’s ChatGPT large language model has proved to be a paradigm shifting event -- with governments and organizations around the world trying to quickly craft safeguards while implementing the technology at a lightning pace.

Finding the Right ‘Guides’

Dante needed guides to help him on his journey through Hell. In the modern AI landscape, Iannopollo said, business leaders will need their own guides. “Welcome to the nine circles of Hell,” she told the audience. “This is the place where AI hallucinates, where AI produces bias and discrimination … Hell is the place where IP abuse and privacy violations and security breaches happen.”

Related:How to Choose a Qualified AI Adviser

Iannopollo said the guides assisting in AI Hell could come from IT, marketing, or the executive team. “All of them understand the incredible opportunity of generative AI and the unparalleled transformative power of the new technology. And they know that without adequate security, privacy, and risk governance.”

According to Forrester’s research, 36% of respondents in those groups said privacy and security are the greatest barriers to generative AI adoption, while another 31% said governance and risk were the biggest hurdle. Another 61% cited concerns that GenAI could violate privacy and data protection laws like the EU’s GDPR.

“So, concerns exist,” she said. “But remember, Hell is a place of confusion.”

As more frameworks come online -- more regulations, there may be less confusion and the guides will help businesses assess their AI adoption. “They may be our starting point here to really think in terms of starting to identify these AI and generative AI use cases.”

A Pit Stop in Purgatory

Once you are out of AI Hell, like Dante, your story is not complete. Dante had to first stop in purgatory. And after spending time in AI Hell dealing with the questions of risk and threats, businesses will need to figure out a compliance strategy.

Related:Biden Pens Landmark AI Executive Order

Iannopollo contends that AI purgatory can be defined as a place of punishment and purification. “Isn’t that the perfect definition of compliance? Punishment and purification? Purgatory is not as scary as Hell, but it’s still a pretty scary place.”

And for this scary place of compliance purgatory, you will need a guide as well, she said. “Who’s gonna be the one to take you to purgatory? Maybe it’s your chief data officer.”

The study showed that 37% of businesses said they were updating existing policies to include generative AI, while 64% said they were working on implementing a brand-new policy. Another 30% said they had no plans for generative AI. New laws impacting data privacy and AI are accumulating both abroad and in the US.

“While we might feel that it is too early to start thinking about specific requirements, I can tell you that a lot of our clients are already working on putting together the team that is going to work on delivering compliance with these new requirements,” Iannopollo said. “If you want to start to think about your dream team, you need an executive champion -- maybe that’s the chief data officer, maybe that’s your CEO, maybe that’s yourself. But you need an executive champion … you want them to be setting the tone of the initiative, this is someone that will own your compliance roadmap.

Related:How Will AI Change the CISO Role?

Empathy and Paradise

Iannopollo cited a study showing that when asked what attributes consumers looked for in brands and organizations, empathy was the trait that consistently ranked high on the list. “And none of the drafts of the regulations or guidelines out there tell you there is a focus on empathy that is going to help us deliver trustworthy AI.

She equates “paradise” to delivering trustworthy AI. Who will be the partners in delivering AI in this paradise? “If you have a chief trust officer, go to them, go to IT, HR … those are some of the experts you need in your community for trustworthy AI.”

And to get consumers or clients to buy into your AI offerings, you’ll need to use empathy, she contends. “Empathy is the ability to understand consumers, employees, citizens’ feelings and needs, and being able to act upon those needs.”

She added, “It’s a long journey that we have in front of us.”