Who’s Afraid of the Big Bad AI (and Data) Wolf?

It’s difficult in today’s business environment to escape being asked about how you are leveraging artificial intelligence (AI). Customers are asking if you are using AI, management is asking how the company can be more efficient with AI or worrying about how competitors are passing you by using AI, employees are asking when they can use AI to help them do their job and more. 

At the same time, these same people have a range of concerns when it comes to AI. So, how do you navigate all of these questions given the constantly changing landscape of AI? Let’s discuss a few of the categories where there is AI anxiety and how you might tackle them:

1. Privacy

One of the most common areas of concern is the use of customer data in AI. Companies have amassed large volumes of data that could be used to drive automation and insights by AI but their customers are concerned about how their data is being used by third parties.

How to tackle: be transparent.

First and foremost, be sure that your company has a privacy policy in place so your employees and customers understand how customer data must be managed and protected. With landmark legislation passed to protect customer data (CPRA, The California Privacy Rights Act of 2020 and GDPR, General Data Protection Regulation), there are increasing requirements for your company to outline how they are safeguarding customer data.

In addition, clearly outline your data policies in customer contracts and agreements even before you have ideas of how you might want to leverage their data in AI and automation. By being transparent with customers and offering them the opportunity to opt out, you can allay much of the concerns surrounding privacy.

2. Bias

AI uses data to automate and provide insights but the underlying data used to train AI can be biased. For example, the news has been buzzing about how AI added to talent recruitment systems was using past recruiting data that was biased towards hiring white males. As a result, the AI recommended white male candidates more often to hiring managers. 

How to tackle: use humans in the loop to check for bias.

Using AI doesn’t mean you shouldn’t check what AI is doing. Be sure that you have checks and balances in place to check that what your AI-enabled tools are doing is free of bias and discrimination. Many companies are leveraging humans in the loop of the training of AI for this purpose. For those concerned about bias, they may feel reassured that humans are involved in the initial training of your AI, even if they aren’t checking every single answer or result.

3. Lack of Confidence and Conviction in AI

For many, the jury is still out on whether AI will be truly transformational or fizzle out and not live up to the hype like other technologies have. So, they are concerned about making big investments now even though they are being pressured or asked to.  

How to tackle: start internal first.

Before you start overhauling all of your systems and processes with AI, consider leveraging it internally first so your company and your employees can get experience with it in less public ways. For example, marketing and sales teams can leverage AI to generate first drafts of marketing and sales content and email subject lines, product managers can use AI to conduct research on competitors or new areas of focus. By starting with internal applications and establishing policies for AI use at your company, you can then explore how AI might be best used for external applications for your customers with more experience and understanding as well as less risk.

4. Ethical Concerns

As AI is being applied to automate more decision-making and recommendations, it raises questions about how we ensure that those decisions and recommendations are being made in a fair, transparent and ethical manner?

How to tackle: build guardrails, be transparent.

What if your customer asks your AI-enabled support chatbot inappropriate questions that are not within your scope or expertise? Work with an AI expert to build guardrails so the AI doesn’t go beyond the scope you would if one of your employees received the question. When you receive an inappropriate question, ensure your AI reiterates your terms of service and scope of service so they understand what is appropriate to ask or what not to ask.

5. Loss of Brand, Human Connection

As AI becomes more prevalent, many companies worry it will remove all personality or human touch in customer interactions. For example, in marketing and customer service, people worry that AI will remove the unique brand voice or service style that differentiates their company.  

How to tackle: train your AI with humans and give an option to interact with a human.

AI doesn’t mean that humans are no longer involved in your brand or service delivery. Automation is intended to do tasks that humans used to do, hopefully in a way that mimics how  humans would do it. You can train your AI to have your voice and personality and your employees can ensure that the AI does what they would have done or said. Also, offering an option to contact an employee can reassure customers that they have a way to a human connection, if needed.  

In Summary:

• Privacy: Be transparent.
• Bias: Use humans in the loop to check for bias.
• Lack of confidence or conviction in AI: Start internal first.
• Ethical concerns: Build guardrails and be transparent.
• Loss of brand, human connection: Train your AI with humans and give an option to interact with a human.

Companies can overcome concerns and fears about AI by taking some key first steps as outlined in this post. The key to new technology adoption is to meet people where they are and help them on their journey whether they have a short or long way to go to adopt the technology.

Featured image by Ray Hennessy on Unsplash

About Synaptiq

Synaptiq is an AI and data science consultancy based in Portland, Oregon. We collaborate with our clients to develop human-centered products and solutions. We uphold a strong commitment to ethics and innovation. 

Contact us if you have a problem to solve, a process to refine, or a question to ask.

You can learn more about our story through our past projects, blog, or podcast.