From streamlining healthcare to helping marketers deliver more compelling copy, artificial intelligence (AI)is transforming the modern world at a rapid rate. When a technology is developed and disseminated this quickly, however, growing pains are bound to ensue.
Our current laws don’t yet have provisions for the unique situations AI will pose. Then, of course, there are customer service concerns, cybersecurity threats, and even the potential for economic stability to worry about.
As Melbourne-based IT services providers, part of our job is to make the transition to new and helpful technologies smoother for our clients. This includes helping them understand not only the benefits of tools like AI but also any drawbacks, limitations, and risks they may present.
Whether you’ve adopted AI, are considering experimenting with some tools, or are simply AI-curious, it’s worth having a solid understanding of the following emerging risks associated with the technology.
Potential for revealing sensitive data
If you’re considering using “generative” AI tools like ChatGPT to create internal documents or customer-facing content, it’s crucial to understand that any information you feed into the AI has the potential to go public. These public AI platforms make it clear in their terms and conditions that they use the data you give them to improve their service. So they may draw on any sensitive information you give them when providing output to other users.
In addition to revealing your proprietary information, this could place you in breach of contract with clients and cause all sorts of other legal headaches. It also leads us neatly on to the second troubling AI risk for business owners.
Difficulty enforcing AI rules in the workplace
You may take the threat to your sensitive data seriously and avoid using public AI tools for drafting contracts or analyzing company data. And if you communicate with your team and offer them adequate training, you can get them on the same page. However, all it takes is one person thinking “surely it won’t matter if I do it just this one time” for confidential information to be compromised.
Setting aside the risk to your data, the use of AI can impact your business in a variety of ways (which we will discuss in the upcoming sections). If you’re not aware that a team member has been using AI, you’ll be completely unequipped to mitigate any potential risks. Of course, you can’t start micromanaging every action every employee takes. However, it does pay to be clear on your expectations surrounding AI, including what tasks can and can’t be completed with it.
Intellectual property laws have yet to catch up
A clear decision has yet to be made regarding who owns the intellectual property generated by AI tools. So any text, images, videos, or other content you create could present a legal issue down the track.
The last thing you want is to be embroiled in a costly lawsuit that damages your reputation. And although AI-related IP laws have not yet been fully solidified, this is a genuine risk with AI-generated content. Precedents are already being set in AI and IP court cases across the globe, so it’s not something to take lightly.
Generally speaking, it remains best practice to respect the rights of artists and content creators. If you’re uncertain about whether you have the right to use certain content, the best course of action is to seek legal advice.
Delivery of misinformation
Whether you’re doing research via ChatGPT or deploying a chatbot on your company website, there’s the potential for the AI to “hallucinate,” generating information that’s convincing but thoroughly false. Hallucinations can arise from coding errors, conflicts in the source data, and biases that emerge as a result of the way the AI is trained.
The last thing you want is to deliver misleading or inaccurate information to your website visitors. So, if you’ll be using AI to generate content or talk to your customers, it’s crucial to take steps to mitigate this risk.
Uncertainty and disengagement in the workforce
AI’s ability to take over repetitive tasks and rule-based tasks has already led to job displacement for many people across the globe. However, sources are now suggesting that everyone from writers and programmers to accountants and psychologists is at risk of being replaced by AI.
No one knows for sure how disruptive this technology will be. However, as the media reports more frequently on AI-related job losses, many people are beginning to feel that their livelihoods are in jeopardy. As you can imagine, this state of uncertainty isn’t conducive to a happy and productive workforce.
To avoid disengagement, it’s crucial to maintain open lines of communication with staff. AI may indeed impact their jobs and your industry. However, you can support employees by keeping them in the loop on any changes and offering opportunities for them to reskill, retrain, and adapt to new roles that leverage AI.
Alienating customers
After the recent Michigan State University shooting, Vanderbilt University issued a statement to students that was supposed to offer some solace to its students. However, the school used ChatGPT to compose the letter, creating a dystopian, dehumanising impact on the recipients. After a barrage of complaints, the school issued an apology.
On a similar but related note, when the bulk of a company’s customer service is taken over by chatbots, it can feel incredibly alienating to customers. Even if you have a customer service or IT support team, customers can get frustrated if they feel like the human connection is being gate-kept by AI.
Gone are the days when you could simply click the Contact Us page and see a phone number and email address. Nowadays, customers often have to navigate through a maze of FAQs or automated chatbot responses before they can finally talk to a person capable of answering their questions.
Of course, this doesn’t mean you should forego chatbots altogether – they can be useful tools, and most customers appreciate them for quick, simple questions. Just take the time to test your bots and get audience feedback to ensure you’re not inadvertently creating an alienating experience.
Creating market uncertainty
Free as they are from emotions and human biases, AI algorithms can be helpful tools for investors who want to make smarter decisions. However, their lack of understanding of human emotion can actually be a hindrance.
Trust and fear play a significant role in market movements, and trading algorithms may fail to account for the impact their actions might have on the emotions of human traders. Given the rapid rate at which trading bots can work, it’s possible for them to create flash crashes and general market volatility.
A single software error at Knight Capital created a US$440 million fallout for the company in 2012, and there’s every reason to expect similar situations to arise from the use of AI trading bots. This is crucial for business owners to understand as your company’s market cap or your personal investment portfolio could be affected.
Potential for discriminatory hiring practices
AI has been floated as a way of getting around discriminatory hiring practices by having an impartial algorithm evaluate candidate applications. However, the problem is that the AI needs to be trained on something, and any data set you train it on will have been developed by humans. So, it comes complete with any conscious or unconscious biases that were at play for those humans.
Even after instructing AI to ignore things like skin colour or disability status in making candidate recommendations, the algorithm often finds a way to select for the biases of the original data set. So, discriminatory hiring practices can be baked in from the start.
These are just a few of the most critical emerging AI risks Australian business owners should be aware of. If the information has raised any questions for you, or if you would like to perform a general IT risk assessment on your business, contact Invotec today. As a managed service provider, we offer a range of bespoke IT packages designed to suit each client’s needs, budget and industry.
