The AI Outsourcing Conundrum

March 18, 2024

It is clear that for outsourcing providers and customers, AI already offers and will continue to offer many ways to improve efficiencies and therefore hopefully profitability for businesses. For customers of outsourcing providers, outsourcing is an attractive option, as it offers the option of accessing innovative AI solutions and expertise in using them. Using AI is not necessarily something all customers can do for themselves and so it offers exciting ways of working for many businesses. But whilst exciting, as with anything innovative, there does come some risk.  

So, what are these potential risks?

It is of course the case that outsourcing will normally rely on technology to some degree. This is not anything new and it is the case that both customers and suppliers will want to address and mitigate any known and foreseeable risks in respect of certain issues - ownership of IP, data protection risk and liability - to name a few.  

However, there are some more distinct issues which particular thought should be given to when thinking of either supplying or using outsourcing services, to try and address the impact of specific AI related risks as follows:

* Personal Data

You’d expect risk in respect of processing and protection of data to be covered in any agreement, but processing personal data using AI may require additional thought and assessment in respect of risk. This would especially be the case if individuals might not expect their data to be processed using AI. We recommend carrying out thorough Data Protection Impact Assessments when using any new outsourcing supplier.  

* Accuracy and Reliability

As with any kind of service, AI poses risk if it does not produce accurate output for a customer. So, both customers and suppliers will want to know if outputs can be trusted and if not, the extent to which the suppler is provided to accept liability for any errors caused by AI. This liability position should be agreed and addressed at the outset by all parties.  

* Intellectual Property Ownership

It will be important to address what party owns what IP and deliverables in the outsourcing context when AI is being used. We have seen that ownership of IP is hotly debated when it comes to using AI but the important point to remember is that the parties should try and agree and foresee ownership issues as much as possible and that any agreed position is put in writing.  

* Bias and Discrimination

It may be the case that AI is used to assist in decision-making processes relating to individuals. As a result, there is an increased risk that the AI may exhibit bias. It could be the case that this means that the Equality Act 2010 is inadvertently breached, for example in respect of its provisions prohibiting discrimination in the supply of goods or services. If either a customer or supplier foresees this as a particular risk area depending on the nature of the outsourcing services being provided, then it needs to be addressed.  

What can businesses do to mitigate these risks?

Currently in the UK there is no legislation in respect of AI. So, customers and suppliers should be seeking to protect their contracting position as much as possible in accordance with existing laws, particularly in respect of data protection and IP.  

We are waiting to see what the outcome of the AI White Paper consultation will bring in respect of regulatory guidance for AI at the end of April 2024 from bodies such as the ICO and the FCA. Watch this space for how it may affect contracting when it comes to AI!

Receive our insights directly to your inbox by signing up to our newsletter

Recommended content