Can customers learn to trust AI?

[HPE_Experts]

AI can be transformative or destructive, depending on the framework it’s conceptualized and used within. Applied to AI, ethics helps us lean into positive transformation, where the technology is a benefit not a hindrance to people.

Do you remember Isaac Asimov? A biochemist and prolific sci-fi writer, his works inspired the 2004 blockbuster I, Robot. It was Asimov who created the zeroth law, which states that “a robot may not harm humanity, or, by inaction, allow humanity to come to harm.” This rule replaced his earlier “three laws of robotics,” an ethical framework he devised in a sci-fi short story to help guide interaction between humans and robots.

Why are we talking about this now? Because while the concept was first introduced in 1942 in the realm of science-fiction, it points to our historic ethical concerns around technology — concerns so pervasive that they made their way into mainstream works of fiction.

Fast forward 80-odd years, and these concerns haven’t dissipated. On the contrary, they have only amplified with the rise of AI.

What are AI ethics?

When we talk about ethics, we’re looking at the moral ramifications of artificial intelligence on human life. We’re asking what it means to make decisions in a world where we’re surrounded by machines that think and act like humans. It helps us set boundaries for what we consider irresponsible or unfair use. Ultimately, it’s about ensuring that AI benefits all people equally, advancing and improving the way we live and work.

Let’s start off on a reassuring foot: AI is a powerful, transformative technology that can amplify human capabilities and open numerous opportunities. But here’s the caveat: it can only be positively transformative if it is designed, deployed, and used ethically.

Why should organizations care about AI ethics?

Earlier this year, HPE conducted a survey of 2,400 IT leaders across 14 markets to understand how and why they are using AI. Within this, one of the things we wanted to know is whether organizations are being mindful of ethics in their pursuit of AI. The answer? Not really. We found that ethics was not only low on the budget-priority list (only 24% incorporated it into their AI investments), but that it was also a significant blind spot in their AI strategy development. That’s because only 11% of IT leaders we surveyed believe ethics is critical to AI success.

Is there good reason for organizations to bump ethics up to the top of their agenda? Absolutely. It’s not only socially responsible but ignoring ethics could have negative business impact. Here, we direct you back to that historic concern over technology. For companies who use AI, ethics will be important for maintaining trust with customers. No trust, no sale is a business reality. Consumer feedback reiterates the link between reputation and use of AI: 64% say they are more likely to see corporates in a good light if these businesses prioritize the responsible use of technology like AI.

Beyond reputational damage, businesses without an AI ethics policy also risk developing models that lack proper compliance and diversity standards, which could bias the output and therefore ROI of the model itself—and open the organization to costly fines and legal battles further down the line.

Following five pillars for ethical AI

Increasingly, consumers expect organizations that are using AI to have governance rules in place to help mitigate ethical issues in a customer-centric way. Not only that, but in an ecosystem of partners, they also expect businesses to check that their ethical values are upheld across the entire supply chain.

As a B2B service provider of AI solutions, HPE is serious about harnessing AI Ethics for Good. We weave ethics into the very DNA of our offerings, guided by five pillars:

1. Privacy-enabled security
   AI solutions must respect individuals’ privacy, be secure, and minimize the risk of errors and unintended, malicious use of the technology.
2. Focused on people
   Respecting human rights is key, meaning AI solutions must be designed with mechanisms and safeguards in place, such as human oversight to prevent misuse.
3. Always inclusive
   Inclusive AI looks at minimizing harmful bias, ensuring fair and equal treatment, and enabling access for individuals.
4. Putting responsibility first
   AI solutions must be designed in a way that supports responsible and accountable use, allowing an understanding of AI and for outcomes to be challenged.
5. Robustly crafted
   Quality testing must be built into AI solutions, which must be engineered to include safeguards that maintain functionality and minimize misuse and impact of failure.

Applied in every instance in the AI lifecycle, these pillars — alongside an internal risk assessment process and employee-based AI ethics training — help us build the trust that is so critical to good business and to using our business to do good.

Ethical AI, in practice

Trust is all-important. At HPE, we focus on delivering AI that our enterprise customers can put their faith in, knowing that accuracy, equity, sustainability, privacy, compliance, and confidence are engineered into our solutions from edge to cloud.

To see how our AI technology creates efficiencies and improves overall output, check out our infographic here. Or, to learn more about how to plan your end-to-end AI strategy for future success that incorporates ethical guidelines, visit our AI solutions page.. 

HPE Experts
Hewlett Packard Enterprise

twitter.com/hpe
linkedin.com/company/hewlett-packard-enterprise
hpe.com