While AI has the potential to increase efficiency and productivity in the workplace, it also comes with legal and compliance risks that must be taken into consideration.
The United Kingdom does not currently have specific AI legislation in place, meaning that businesses are at risk from both inputs to and outputs from AI systems, and the unknowns the future may bring.
What business risks are associated with AI?
Inaccurate information: AI repeats what it has been told, or what it finds online, it does not actually ‘know’ anything, meaning it may not be telling you true or factual information- Data Breaches: uploading confidential or private to AI creates a potentially uncontrollable security risk, particularly when used by people who don’t know the risks.
- Bias and Discrimination: as above, AI is taught, which means the information it provides will reflect its ‘teachers’ bias’s, putting a business at risk.
- No originality: AI cannot think for itself, it can’t have new ideas. If used by businesses in this way, you have to consider where the idea has come from
- Overreliance: If AI is used to perform multiple roles in a business, who in the business is trained to carry out the job if things go wrong; how do you know if you are providing the correct information to our employees or customers if you are not trained in the information the AI is providing.
- Legal and Regulatory errors: If AI provides you incorrect legal or even safety advice and something does wrong, who is liable?
- Commercial distrust: If a customer recognises your product is being advertised by AI, how, in a world of scam artists, can they believe the product is real or as described.
Is AI considered intellectual property theft?
Currently the unfortunate answer is maybe but there are no clear guidelines being set yet in the UK.
You must remember when using AI, it can only regurgitate information it has been given.
If you think about the AI posters popping up all over social media or shop windows – they all look the same, and they have all copied that design from somewhere.
If AI creates something, be it a policy or a poster, that is discovered to be a copy of someone else’s intellectual property, businesses could get caught out in future.
Can you trust AI to be accurate?
No, and more controlled AI channels have started to tell users this too, so hopefully they are listened to!
If you ask AI a question, you must check that the answer provided is current, accurate to the country you are in, is unbiased, based on fact and within the correct context.
AI has also regularly shown it will make something up to satisfy the user as it has been taught to ‘say yes’!
All this means that though it can absolutely help provide support, you do have to be knowledgeable about the subject you have asked about.
How does AI affect confidentiality?
If I were to, for example, ask an AI program to provide a report on our employees wages and demographics, I have to provide the information to the program.
Do I have permission to upload my employees personal details? Do I know what security the AI has in place to protect this information? If someone else asks the program about company wages, will it provide the details to them?
You just don’t know; so shouldn’t risk it unless you can confidently answer these questions.
How can WA Management help?
WA Management offer an Understanding AI online training course, suitable for staff at all levels across all industries that addresses global best practice around AI use.
Cyber Security and Understanding AI training courses are essential tools in keeping your business safe from cyber security threats and AI misuse. Make sure you don’t miss out on our 10% off deal on these courses, available until the end of October. Simply enter the code ‘ai10’ at checkout to save!
