Five things to do before going all-in on AI

By Natalie Wright, managing partner at the Leeds office of  Forvis Mazars

Artificial intelligence (AI) is a technology that has not only captured the attention of most industries, but has also sparked widespread discussion among businesses and the general public alike. Privately owned businesses are no exception, they are increasingly exploring the opportunities AI presents, while also grappling with the ongoing concerns of cyber threats.

Business owners are currently pursuing how they can harness the power of AI to achieve their goals – whether that’s increasing employee productivity, reducing costs, simplifying operations, improving their security posture, retain a competitive advantage, or enhancing the customer experience. They are working through practicalities as they transition from AI ideation to implementation.

With such enormous potential, it can be tempting to green light a major investment in AI with the view that the sooner you start, the faster you’ll start reaping the benefits, especially if your competitors are already using AI. However, without having the right foundations in place and a clear strategy for what you intend to achieve, the implementation of AI can prove to be a costly mistake. It is critical that as a business owner you understand where risks reside before embarking in your AI journey.

“It’s important to get back to basics,” says Asam Malik, partner, technology & digital consulting at Forvis Mazars.
“Do you understand the risks? Do you understand the opportunities? Can you identify the relevant regulations? Are you producing a business case or a return on investment (ROI) analysis on AI? Because if you get AI wrong, rather than benefitting from artificial intelligence, you’re just left with artificial information.”

Here, we list five key foundational steps that should be in place before going all-in on AI to unlock its potential.

1. AI cannot be fully leveraged without a Digital Strategy
AI is a disruptive technology, and to handle it properly, it is necessary to implement a digital strategy that integrates AI into the rest of the technology landscape within the business. As part of that, business owners and leaders need to facilitate the adoption of AI, not restrict it. They need to think about how to disrupt their own business. At the same time, before AI is implemented the use cases for AI must be identified and pinned down.

2. AI is only as good as the data that it is fed
For AI to be effective, the underlying systems within the business need to be stable and the data needs to be accurate and comprehensive, otherwise it will be the old IT adage of ‘Garbage in, Garbage out.’ “AI has become very accessible,” says Malik. “It’s easier to get AI tools and products now than it ever has been. But the underlying data that AI is pulling information from is still the same. And if that data isn’t accurate, or not comprehensive, you’re creating more of a risk because you’re making judgments off the back of incomplete or inaccurate information.

“The key thing with AI is having confidence in your data and making sure it’s comprehensive. You can’t miss that foundational step.”

3. The security risks around AI in particular cyber are often overlooked
Criminals are leveraging AI to launch sophisticated cyberattacks. For example, they’re using AI to craft convincing phishing emails and even acting as the company’s CEO in a deepfake video conference. So it’s just as important to put the right safeguards in place when you’re implementing AI. “After implementing AI, one client gave it free rein across the entire network – which you should do to leverage the benefits of AI as you need to get access to as much data as possible. But you also need to lock down the data that you don’t want AI to have access to,” says Malik. “At this particular company, someone tapped into the AI system to find out confidential information about employees’ finances and HR records. So you’ve got to be careful around putting the safeguards in place around it.”

4. The data privacy risks around AI are not always considered
Data privacy laws and regulations are about making sure that the data is only used for the purpose it has been gathered for, and permission is obtained from the individuals involved to do so. But there is the potential with AI to use that data for analysis around much broader areas than for which it was originally collected. That, of course, can lead to regulatory fines, reputational damage and a loss of confidence from customers, employees and investors.

5. Skills – it’s a two-fold problem. There is both a significant gap in AI skills in the market, and there needs to be better AI literacy for employees

The shortage of talent required for innovation is a global issue, but it’s felt more acutely in the UK, with 72% of IT and business decision-makers acknowledging this gap, it becomes an even greater issue when you look at privately owned businesses in the UK. This underlines the urgent need for learning agility, AI fluency, and creative thinking.

“Similarly, you can implement any technology across a business, but if you don’t provide the training in the first place, employees won’t know how to use it or be able to leverage its benefits,” says Malik. The above steps may seem obvious, but they are often overlooked. And the only way to successfully leverage the power of AI for privately owned business is to have the proper foundation in place first. Like most things, the basics shouldn’t be overlooked.

For further information on how our technology and digital experts can help you click here