Support 110 years of independent journalism.

Advertorial: in association with Linkedin

How AI can help unleash employee potential

With careful regulation, AI can move us to an economy truly based on our skills.

By Sue Duke

LinkedIn has been using AI for years to advance its vision of creating economic opportunity. With over 930 million members, 70 million businesses and upwards of 15 million job openings published, it is a task beyond the human operator to match jobs to potential candidates. We use AI to build tools to help jobseekers find their next role, and for recruiters to discover their next hire, better matching their skills to needs. We strive to do it in a responsible, transparent and accountable way.

On a macro level, using AI in recruitment brings massive benefits to the economy. Recent research from Oxera found that using AI tools in recruitment generated $168bn in the USA, €104bn in the EU, and £22bn in the UK in 2019, or 0.79 per cent, 0.74 per cent and 0.98 per cent of GDP in the US, EU and UK respectively.

We have a labour market that has long been opaque, inefficient, and inequitable. And we really are starting to hit these constraints now. The shortages of skills and talents in particular roles and sectors in the UK are having an adverse impact on those businesses and on the economy as a whole. In the UK, skills-first hiring increases the candidate pools of eligible workers who were previously overlooked by ten times, compared to the more traditional approach of building pools of candidates by matching particular degrees or job titles. AI can help open that up at scale, so that workers are no longer defined by the last job title they had or the particular degree they studied, but the range of skills and experiences they can draw on.

This is particularly relevant to new technology-driven sectors of the economy where there just aren’t the named degree-level qualifications, such as in AI prompt engineering. Even in more established industries such as education, if the hiring approach is limited to candidates with specific degrees, the talent pool might be relatively narrow. However, by adopting a skills-first approach, the pool of potential candidates can significantly widen. On LinkedIn, for example, broadening the search criteria to include skills related to the industry rather than specific qualifications alone could increase the talent pool by up to 14 times in the UK. This approach enables the inclusion of a diverse array of individuals who possess relevant skills, even if they don’t hold a particular degree.

AI can also be used to understand the emerging trends in the labour market in terms of what skills are in demand and inform the training and education of workers in a more dynamic way. LinkedIn works with partners to develop Local Skills Improvement Plans (LSIPs) in UK regions to use our data to identify the skills needs that are emerging, the skills that are most in demand over the coming period, and the existing stock of skills that sit in those labour markets. That means we can support government programmes to identify and quantify the gaps, and we can then target and fund reskilling programmes that are going to give us much more return both for workers and for companies in the economy at large.

Using AI is also a means of tackling some of the inequalities in the labour market through a skills-first approach to recruitment. On average, for workers without degrees, we see a 9 per cent relative increase in their representation in those talent pools. For women, we see not only more coming in, in absolute terms, to those talent pools, but crucially, and especially in roles where women are under-represented, we see 24 per cent of relative boost in their representation. This means not only are your talent pools getting bigger, but they’re bringing in talent that has been historically overlooked. This drives equity, as well as the productivity of the economy.

We have developed best practices and norms since we started using AI over 14 years ago and codified them early this year as the LinkedIn Responsible AI Principles. These principles include accountability, transparency, trust and fairness. As a concrete example, we will work to ensure that our use of AI benefits all members fairly, without causing or amplifying unfair bias.

Beyond our own organisation, we recognise that governments and societies around the world are working to figure out how to make AI serve humanity and to help ensure that it is safe and useful. As best practices and laws around governance and accountability evolve, we will embrace those practices. The UK, for example, is taking a principles-based approach to AI policy and sets out the key norms for the entire life-cycle, from the conception of AI to full deployment.

If we’re going to meet society’s challenges, we will require more people with different talents. Skills are changing so rapidly that if we continue to do things the way we’ve always done them, we won’t be able to keep pace with that change. AI offers us the chance to do that differently, if we use it well and responsibly.

[See also: There is no chance the government will regulate AI]