Outsourcing service providers are increasingly seeking to rely on robotic process automation (RPA) as a way to reduce costs and error rates, improve regulatory compliance, and provide services more efficiently.

RPA is the use of software with machine-learning capabilities and artificial intelligence to manipulate data and perform repetitive, rules-based tasks that were previously performed by service providers' employees. RPA "robots" can perform anything from basic tasks, such as data gathering, to sophisticated analysis using artificial intelligence to understand meaning and context from unstructured data.

Despite the headline-grabbing success stories, little discussion has taken place on the impact of RPA on the terms of outsourcing contracts. Most commentary has focused on the typical benefits of RPA and on advising customers to ensure they receive a reduction in costs when their service provider is using RPA to streamline and improve the services. The adoption and application of RPA does not merely affect cost, however. Other issues include:

  • Who is responsible for mapping the various systems and processes to be replaced by the robots?
  • Who is responsible for developing the decision-making parameters and exceptions to be applied by the robots? Who would be liable if a recommended course of treatment does not result in a successful outcome for a patient? Who would be liable if a mortgage application was incorrectly rejected?
  • How does an RPA-based solution fit within an organisation's enterprise level security? There will understandably be nervousness about the creation of new automated systems and the impact that that will have on security of the enterprise.
  • Will the traditional definition of key employees, such as the account managers, remain applicable? Arguably, the employees who have programmed the robots (both those who write the code and those who "train" the robot to do the right thing) are more important to the success of the project, so both customers and service providers will want to ensure that appropriate processes are in place to ensure that the loss of the employees who program the robots does not adversely affect their continued operation.
  • How does RPA sit with existing value for money mechanisms? Unless contract terms are properly constructed, customers risk relinquishing to service providers the majority of the benefits of RPA advances. Contractual benefits-sharing mechanisms will become increasingly important.
  • How to address "ownership" of the robot/RPA engine? Who owns the IP rights in the software or rules engine, the service provider or the customer? On what terms can a service provider leverage its solution for one particular customer on behalf of other customers?
  • What happens on termination of the contract? Is the service provider entitled to retain the robot/rules engine or does the customer have the right to obtain the source code/rules and prevent the service provider from continuing to use it?
  • How to ensure interoperability? The customer and the service provider will need to ensure that the robots are compatible with the customer's present and likely future systems. In general, the evolution of automation standards will be a key gating factor on the speed of mainstream implementation of RPA.
  • Data collection and privacy. The customer will need to ensure that it has appropriate consents in place from individuals whose personal information is processed by robots and that it imposes appropriate contractual obligations on the service provider with respect to the storage, access and processing of that information.

Will the adoption of RPA result in the early termination of existing outsourcing contracts? Unlike traditional outsourced services, RPA can be deployed in-house if the correct skills are available.  Given that RPA can be programmed to perform specific tasks for a particular business, customers may be less likely to purchase commoditised off-shore outsourcing services may not be "quite right" and that may result in the payment of substantial set-up and customisation costs to the service provider. Customers may also repatriate work from off-shore locations if the level of service provided by the robots is better and more consistent than that currently offered by its service provider.

RPA will remain a key consideration for CIOs in 2016 and beyond. While the information available to-date supports the view that RPA can be very beneficial for certain businesses, companies ought to consider the broader legal, technical, security, cost and staffing implications of any proposed RPA solution. The timing of RPA's evolution from growth stage to maturity depends on many factors, and that includes the degree of pragmatism and realism of the claims that service providers make for the benefits that RPA can deliver.