AI tools and the use of AI to solve all manner of problems/questions/crises has well and truly exploded. There seems to be a “chat” or “bot” for almost everything. While AI can be an incredible tool, there is always a risk of error and although some errors will be of little significance, some, especially in the workplace, could have catastrophic consequences. The widespread use of AI is still relatively new and has enjoyed exponential growth, making it essential, albeit difficult, for businesses to keep up with the management of its use by staff for work-related tasks.
Data protection and the precision of responses
Companies in all manner of sectors are investing in AI to help move their businesses forward and with the right oversight AI, particularly those tools that are built with a specific business or industry in mind, can be incredibly useful. Where issues are perhaps more likely to arise is when staff take it upon themselves to use an external AI chat or bot to carry out their day-to-day tasks without authorisation. For example, asking an AI tool to tell them the answer to a client query or a regulatory issue they may be faced with. While well intentioned, if workers are using AI in this or similar ways without the company’s knowledge, it creates a situation where the employer loses control over how work is carried out. Even though most employees would not have nefarious motivations (save perhaps for trying to cut corners to make their working life a bit easier), the risks of using AI nevertheless remain. In advisory industries such as law or accountancy, this could be particularly dangerous given clients are relying on the advice given to them and, if it later proves to be incorrect, could lead to negligence claims which are time consuming and expensive. A defence of “ChatGPT did it” is unlikely to hold much credence!
The use of AI can also prove problematic from a data protection perspective. If an employee shares client information, personal data or commercially sensitive information with an open third-party AI tool, the data could be stored and used. Should that data go on to be leaked or misused, the company would likely be liable for the breach even if it had no idea the employee had used it in such a way.
Guidance and establishing good habits
Employers should take action to control the use of AI in so far as they can by ensuring they provide clear guidance on what is and isn’t acceptable. A good starting point is a robust policy setting out expectations around the use of AI in the workplace, coupled with training for employees on the application and meaning of that policy. Some employees may not appreciate the potential harm or may think that an AI tool is infallible, and it is important to make sure they realise that this is not the case.
In some circumstances, the unauthorised use of AI in the workplace could be a disciplinary matter and staff should be reminded that disciplinary action may be appropriate regardless of the outcome. Employers may wish to ask their IT team to block access to certain tools on work devices to add an additional barrier to their use, but bear in mind that this will not solve the problem entirely as staff will be able to bypass any block by using AI on their personal devices, so emphasising what can and cannot be used for work and the consequences for non-compliance is really important.
While the advent of AI is no doubt exciting in many ways, if it isn’t used in the right way and/or its limitations aren’t appreciated it could cause more problems than it solves. Employers need to be mindful to avoid such issues by making sure they have a plan in place to avoid the unsanctioned use of external AI tools in the workplace which will help also avoid the potential headaches and legal liabilities that may come with it!
