Why ethics is becoming AI’s biggest challenge

Many organizations are either delaying or pulling the plug on generative AI due to concerns about its ethics and safety. This is prompting calls to move AI out of technology departments and involve more non-technical business stakeholders in AI design and management.

Also: The most popular AI tools of 2024 (and what that even means)

More than half (56%) of businesses are delaying major investments in generative AI until there is clarity on AI standards and regulations, according to a recent survey from the IBM Institute for Business Value. At least 72% say they are willing to forgo generative AI benefits due to ethical concerns.

More challenging than technology issues

Many of the technical issues associated with artificial intelligence have been resolved, but the hard work surrounding AI ethics is now coming to the forefront. This is proving even more challenging than addressing technology issues.

The challenge for development teams at this stage is “to recognize that creating ethical AI is not strictly a technical problem but a socio-technical problem,” said Phaedra Boinodiris, global leader for trustworthy AI at IBM Consulting, in a recent podcast. This means extending AI oversight beyond IT and data management teams across organizations.

Also: 15 surprising ways I used AI to save me weeks of work in 2024

To build responsibly curated AI models, “you need a team composed of more than just data scientists,” Boinodiris said. “For decades, we’ve been communicating that those who don’t have traditional domain expertise don’t belong in the room. That’s a huge misstep.”

AI teams should ask themselves these questions

It’s also notable that well-curated AI models “are also more accurate models,” she added. To achieve this, “the team designing the model should be multidisciplinary rather than siloed.” The ideal AI team should include “linguistics and philosophy experts, parents, young people, everyday people with different life experiences from different socio-economic backgrounds,” she urged. “The wider the variety, the better.” Team members are needed to weigh in on the following types of questions:

• “Is this AI solving the problem we need it to?”
• “Is this even the right data according to domain experts?”
• “What are the unintended effects of AI?”
• “How can we mitigate those effects?”

Very important strategically

Business leaders may be growing more cautious about the ethical implications of AI, but they also see a strong embrace of ethics as a source of competitive strength. Seventy-five percent of executives view AI ethics as an important source of competitive differentiation, and a majority — 54% — expect AI ethics to be “very important strategically.” It’s an important signal to stakeholders: more than 85% of surveyed consumers, citizens, and employees value AI ethics.

Also: I’m an AI tools expert, and these are the only two I pay for

A holistic AI ethics framework identifies three types of ROI that can result from AI ethics investments, the IBM report states:

• Economic impact (tangible ROI): The tangible or direct financial benefits of AI ethics cover measurable factors “such as cost savings, increased revenue, or reduced cost of capital. For example, an organization might avoid regulatory fines by investing in AI risk management.”
• Capabilities impact (long-term ROI): This alludes to the long-term benefits of an AI ethics effort. Examples may include “technical infrastructure or specific platforms for ethics may allow organizations to modernize in ways that lead to further cost savings and innovation.”
• Reputational impact (intangible ROI): The intangible or difficult-to-quantify benefits coming out of a strong AI ethics effort cover factors such as brand and culture that positively affect “an organization’s reputations with shareholders, governments, employees, and customers.” This may include “improved environmental, social, and governance scores; increased employee retention; and positive media coverage.”

Many executives and managers are not fully tuned into the three potential areas of impact that AI ethics efforts can deliver, which requires an ongoing education process. Boinodiris advises engaging “your savviest AI ethics experts to educate the C-suite on differences between loss aversion and value generation approaches to AI ethics. Help executives envision the potential of leveraging AI ethics technology, platforms, and infrastructure for broader use.”