AI is making it’s stance within the recruitment industry, with more and more recruiters and employers choosing to stream their hiring process through the technology. In this blog, we’ll unpack the ethical questions surrounding AI and challenge the role it plays in today’s business landscape.
Here’s what we’ll be diving into:
- What happens to the human touch in an AI-led hiring process
- How bias can still exist, and even grows through AI-assisted decisions
- The growing concerns around data security and misuse in an AI-driven world
When attending the Recruitment Expo in October 2024, it was very apparent AI was the common theme, with many seminars talking about the integration of AI within recruitment and the narrative that ‘AI will not replace your jobs but recruiters using AI will’.
Researchers have found that by automating repetitive, administrative tasks it has significantly boosted efficiency and productivity, while allowing recruiters to prioritise value-added services like relationship and culture building during hiring. Other researchers state AI in recruitment has emerged as a significant trend because this technology can process data and make decisions at volumes and speeds far surpassing human capabilities.
But before you rush in to start applying AI into your hiring process, here are some ethical factors to consider…
A recent SSR study revealed that 70% of enterprise-sized businesses in the UK now use applicant tracking system (ATS) software to screen CVs. However it’s important to note that they can unintentionally reproduce and perpetuate diversity biases, discriminating against candidates based on personal characteristics. Amazon scrapped an AI recruitment tool after they found the algorithm was favouring male applicants and penalising CV’s from female applicants. Amazon found
that the system had taught itself that male candidate were preferable.
Critics also argue that AI cannot replicate the emotional intelligence and cultural insight that human recruiters bring to evaluating applicants. Over-reliance on automated systems may diminish the nuanced judgement needed to fully assess candidates, especially since many ATS platforms screen CVs based solely on keywords, automatically rejecting up to 75% of applicants. A study found that applicants are more likely to apply for jobs using human rather than AI-aided evaluation, as they prefer to maintain a human element.
In an evolving world where AI and automation are transforming the way we interact, maintaining a human touch is not only a competitive advantage but a customer expectation.
Studies show consumers actively trying to bypass automated systems in favour of speaking with a real person. Enterprise Times, found in a study conducted by ServiceNow:
- 75% of UK consumers press no buttons during automated chatbot-enabled telephone calls to be redirected to a human as quickly as possible
- 77% of UK comsumers are looking up tips and “hacks” to avoid automated customer service conversations and talk direct to a human
While AI-powered chatbots can improve efficiency, they often lack empathy and a personal connection – which are elements that matter to customers. In recruitment, where trust plays a central role, balancing technology with authentic human interaction which is essential for building relationships.
Assessing candidates the human way allows recruiters to evaluate qualities that AI often misses, such as soft skills, cultural fit and adaptability – factors that are especially critical during a skills shortage. While AI can efficiently screen for technical experience, human judgement is essential for understanding a candidate’s communication style, problem-solving approach and potential to thrive within a team, ensuring that hires are not just qualified on paper but also aligned with the organisation’s needs and culture.
A vast amount of research suggests that employees are often left to navigate the use of generative AI tools without clear organisational guidance or policies, finding that when no company guidelines are available, employees attempted to manage these risks themselves, terming this adoption as ‘self-guided strategies’. Without formal policy or guidance it presents potential risks, including employee misuse, breaches of data privacy and violations of the General Data Protection Regulation (GDPR),.
According to a Statista and Semrush report, the UK ranks among the top 20 countries for desktop and mobile traffic to ChatGPT between July 2023 – April 2024, with full-time employees accounting for 37% of users. Statista and Semrush also found that 75% of the respondents highlighted writing work communications and automating tasks as key functions, with organisations worldwide adopting generative AI for the following anticipated benefits:
- Improve efficiency and productivity (56%)
- Encourage innovation and growth (29%)
- Shift workers from lower to higher value tasks (26%)
A global data collection revealed that, between 9th – 15th April 2023 there were 319 incidents of sensitive data leakage per 100,000 employees on ChatGPT.
The National Cyber Security Centre state that, unless a user opts out, OpenAI’s usage policies specify that user inputs may be stored and used to further train the model.
Together, these findings stress the importance of providing organisational guidance. While employees may show initiative in attempting to protect data, their self-determined strategies are inconsistent. If formal guidance and training are not provided, the risk that employees could unknowingly misuse ChatGPT increases, either by them misunderstanding how data is stored and processed or by overestimating the effectiveness of their self-guided strategies. As a result, providing internal policy and making training mandatory for all employees is essential to minimise risks and support ethical and responsible adoption. as data could be easily included within documentation or drafts submitted for proofreading or refinement and might unintentionally enter an AI model.
Some researchers observed in their study that managers are frequently unaware of how widely ChatGPT is already being used within their teams.
Without organisational or national standards, if a mistake arises, who is liable and bears the responsibility?
Open AI’s terms of use place the responsibility on the user to apply with their applicable laws, including the prohibition of compromising the privacy of others without complying with applicable legal requirements.
Many believe that banning AI is not a realistic solution, as employees could find alternative tools or use the technology in secret. It has been recommended that businesses:
- Offer guidance on anonymising data
- Providing an internal policy on AI use with the organisation
- Clarify how privacy settings can be adjusted
- Where feasible, providing access to private paid versions of AI
- Providing training to employees
Innovation should move businesses forward – not away from people. While AI can streamline processes and improve efficiency, businesses must remain aware of the ethical considerations behind it, including bias, data protection, and maintaining a human-led experience. The future of recruitment and business support should be powered by technology, but guided by people. The most effective approach isn’t choosing between people or technology, but finding the right balance between the two.
The right hire still starts with people. If you need support finding the right talent for your business, get in touch with our team today.
