What will be the impact of new data portability and transparency rules and the proposed "high-risk" regime for use of AI in workforce management and recruitment?

Osborne Clarke, in partnership with the European Company Lawyers Association, has released a report exploring the challenges and opportunities associated with data-driven business models (DDBMs), with which workforce-solutions companies – including platforms and more traditional businesses – are increasingly involved. For example, companies are using their own and customer data sets to support the planned use of artificial intelligence (AI) in processes such as matching candidates to vacancies and using data sets to offer data-related analysis and consultancy services.

The report, which is based on responses from 419 companies in the UK and Europe, highlights that:

• 89% of companies across all sectors now offer or plan to offer data-driven products or services
•  87% of technology, media and telecoms businesses surveyed offer data-driven products and services
• 35% of companies see data as enabling them to offer new products and services, and we suspect the percentage will be higher for workforce solutions companies
• 58% of companies offer or plan to offer data-related analysis and consultancy services, and we suspect the percentage will be higher for workforce solutions companies
• 100% of companies looking to build usage of AI will rely on ownership of the data sets underpinning the AI tools they use

Data-rich activities

Many workforce solutions businesses are further digitising their processes and service offerings and collecting and using huge amounts of data – a lot of which is personal data (but a lot of which is not). The report focusses on opportunities for businesses around the use and development of all types of data and not just personal data and how businesses need to think strategically about their use of data and beyond compliance with data-protection legislation.

The report looks at how regulation for DDBMs is on the cusp of a major expansion and how businesses will need to adapt or design new models to work within the changing legal environment across Europe and in the UK. The increasing complexity and scope of data and digital regulation means that workforce solutions and other companies require new skills and a "hands-on" proactive mentality to thrive in this new world. 

Regulating the previously unregulated

For example, data regulation is poised to expand far beyond privacy and personal data into new, previously unregulated areas. European Union (EU) legislators are seeking to reshape its data ecosystem to open up access to data and to ensure alignment with its values and fundamental rights. The portability of data from one platform to another will be boosted – and this may affect dominant players in particular areas. New EU legislation also proposes regulatory frameworks for the data ecosystem, including for "data intermediaries" that supply and facilitate the availability of data. This raises some interesting questions for the sector: will companies continue to "own" what they think they own now? Will they have access to lots of new data? What can be done to prepare for this?

Opening more data to be available to more businesses is a particular focus of policymakers, because of the recognition that data is a key factor for production. This is particularly the case for machine-learning and deep-learning forms of AI, which are typically developed and honed using huge quantities of data. Access to more data could power new applications of AI, enabling greater efficiency and new insights. 

40% of respondents expect to increase their use of AI in business development (and the percentage will be higher in larger workforce-solutions companies), highlighting the need for data to power the expansion. AI systems considered to be high risk will be subject to extensive regulation with an enforcement framework at national level, which includes the potential for heavy General Data Protection Regulation-style fines for non-compliance – up to 6% of global turnover. AI systems designed to interact with humans will be subject to strict regulation and transparency requirements The high risk AI regime for workforce solutions companies

The high-risk regime will make sure the person concerned knows they are dealing with an AI tool not a human. Rules will cover things like governance and management practices and design choices; data collection; relevant data preparation processes; the assumptions about what the data measures and represents; prior assessment of the availability, quantity, and suitability of the required datasets; examination for possible bias; and the identification of possible gaps or shortcomings in the data and how those gaps and shortcomings can be addressed.

Data sets must be relevant, representative, free of errors, and complete, and must have appropriate statistical properties for the intended application of the AI tool concerned. Where appropriate, datasets must take into account the particular characteristics or elements of the specific geographical, behavioural, or functional setting in which the high-risk AI system will be deployed.

Osborne Clarke comment

The UK seems likely eventually to have a similar regime to the EU and, in any event, larger workforce-solutions companies will need to design systems which cover both areas. Companies should be prepared for their business and clients to require direct contractual relationships with many more parties such as developers and hosting providers.

For some businesses, the data provisions in a draft contract will be the first to be ceded in negotiations; for others, a data-savvy lawyer will sit at the table in every deal. As understanding of how to leverage the value in data increases, this difference in attitude can potentially start to impact on revenue streams and lost opportunities. The company's intellectual property (IP) strategy may need to be expanded to ensure that any IP in data or data-bases is effectively protected.

We have worked on cutting-edge data projects to help others consider the possibilities and help them navigate the complex and developing regulatory landscape. Let us know if you would like to know more about how the use of data is developing and how we can help support this.