Navigating AI Ethics: Why It Matters and How to Build Trust

19 April 2024

By Aki Antman, CEO & Founder of Sulava, President of AI & Copilot at The Digital Neighborhood, Peter Charquero Kestenholz, Founder, Head of Innovation & AI at Projectum and Erik David Johnson, Chief AI Officer at Delegate.

This blog is the first in a series of eight blogs about the challenges and opportunities in AI technology.

AI is transforming the world in unprecedented ways, offering new opportunities and challenges for various sectors and industries. However, as AI becomes more pervasive and powerful, it raises ethical questions and concerns that must be addressed to build and maintain trust in the technology.

Ethical Considerations

Among the foremost ethical considerations in AI is the assurance of fairness, transparency, and accountability in developing and deploying AI models and applications. Given that AI systems can perform significant tasks and have an increasing influence over human lives and well-being, particularly in domains like healthcare, education, and criminal justice, it is imperative to prevent discrimination, harm, or violations of individual or group rights.

Central to this endeavor is ensuring that users and stakeholders comprehend the workings of AI systems, including the data they rely on and the outcomes they generate, enabling them to identify and rectify errors or biases. While some AI models offer explainability, many, particularly those based on deep learning and neural networks, operate as "black boxes," making it challenging to elucidate the reasoning behind their decisions.

Responsible AI practices necessitate caution in deploying such opaque models, urging the exploration of alternatives that offer greater transparency and explainability. Moreover, ethical considerations extend beyond technical aspects to encompass the appropriateness of AI applications, recognizing that certain uses, such as automated decision-making in sensitive areas, demand heightened scrutiny and accountability.

For example, if you are an insurance company and tell a person they cannot get health insurance from you because the AI model says no, you will have a legal problem if the model is not explainable.

Ethical guidelines serve as essential guardrails, guiding the development and deployment of AI systems in alignment with societal values and norms. Embracing best practices, such as assembling diverse, multidisciplinary teams and conducting thorough ethical impact assessments, fosters responsible AI development, prioritizing societal well-being over practical solutions.

And typically, the ethical solution is also a future-proof solution because the ethical solutions we are making right now are future-proof for new legal systems, such as the EU’s new AI regulation, which comes into force this fall.

"Typically, the ethical solution is also a future-proof solution because the ethical solutions we are making right now are future-proof for new legal systems, such as the EU’s new AI regulation, which comes into force this fall."

Erik David Johnson
CAIO, Delegate

Trust in AI is growing

Another key issue is that building and maintaining trust in AI technology is paramount for widespread adoption and acceptance, influencing users' perceptions, behaviors, and satisfaction levels. While AI holds immense potential, particularly with advancements in Generative AI, its integration must be underpinned by a deeply responsible approach emphasizing knowledge and experience.

Trust hinges on several factors, including AI systems' reliability, accuracy, safety, and their alignment with user values and expectations. Effective communication and interaction between AI systems and users further bolster trust, fostering transparency and user confidence in the technology.

Right now, we are in the age of "large language models," which for the last 18 months has led to a massive increase in daily AI users, as programming skills are not needed. This allows users to experiment on their own and, for some, build up more trust in how it works.

And trust in AI has grown, with a notable shift from skepticism to increased awareness and acceptance, as evidenced by rising adoption rates across various regions.

According to a newly released report, the use of AI in the Nordics is largely viewed positively, especially in Finland. 61% of Finnish organizations use AI, compared to 52% in Norway, 48% in Denmark, and 45% in Sweden.

Still, European countries are far behind the US, and we need to accelerate the pace to stay caught up.

As companies become more aware of different AI models available in the market and how they work under the hood, they adopt a more welcoming attitude toward AI. One approach is to enforce AI top-down, but often, a bottom-up approach will build trust in AI technology, as individuals experience benefits firsthand before broader company-wide implementation.

Yes, upper management needs to understand AI in order to make the right strategies. But people on the ground also really need to know what it is to ensure that efforts to utilize generative AI are well received among those who will actually use it. And that's where the real value is.

"The European countries are far behind the US, and we need to accelerate the pace to stay caught up."

Aki Antman
CEO & founder af Sulava, President for AI & Copilot hos The Digital Neighborhood

en_GBEnglish