AI in HR: How leaders can experiment safely to boost productivity

Only 15% of HR teams are implementing artificial intelligence (AI), while the rest are in ‘wait and see’ mode – a hesitation that risks sidelining HR from the strategic role it has long pursued. Yet HR professionals are right to approach AI with justifiable caution. A balanced approach is essential: if HR waits too long to engage it risks missing AI’s transformative benefits. But if it charges ahead recklessly it risks issues with ethics, security and job displacement.

Sarah Franklin, CEO of AI-enhanced people platform Lattice, points out, however, that this is the moment for HR to seize the opportunity of AI.

“There's so much opportunity to change how we learn, how we lead, how we grow,” she says, “but we need you to have the confidence and the courage to say, I don't know all of the answers but I have the courage to ask the questions. 

“Revenue leaders are going to care about the money, marketing about the pipeline, the CIO about the technology. Your ‘North’ is uniquely people. And that is why we need you to help build a future that is people powered, where AI doesn't happen to us, it happens in service of us.“

This article makes a case for HR leaders to begin low-risk, high-value AI experimentation to harness its productivity and creativity-boosting powers. Based on insights from industry voices and real-world examples from Lattice’s Lattiverse event, the following guide aims to give HR professionals practical, manageable ways to integrate AI into their workflows.

Why should HR leaders embrace AI experimentation?

Paul Devoy, CEO of Investors in People, perhaps says it best when he likens AI adoption to a boat leaving the harbour. Those who don’t get on board now will soon find themselves unable to catch up, as AI begins to shape HR in ways that extend far beyond automation. 

“I've been working nearly 40 years and this is potentially the biggest change that I've seen in my entire career,” he says, pointing out that productivity stagnation in the UK – a persistent economic challenge – can be tackled with AI’s help, providing HR departments make strategic use of its strengths. 

“If we want to create better public services, we want to create more well-paid jobs, we want to create a more prosperous society, AI is potentially a massive tool in helping us to fix that,” he says, “so organisations absolutely need to embrace it but it does come with lot of inherent risks.” 

Implementing AI thoughtfully in HR can lead to:

• Increased productivity by automating time-intensive tasks
• Greater fairness and consistency in hiring and performance reviews
• Enhanced support for “accidental managers” through AI-driven guidance.

For those ready to tackle those risks AI presents a rare opportunity to boost efficiency, free up managers’ time and open new possibilities for creative problem-solving. However, success requires a cautious, strategic approach to implementation, starting with low-risk experiments.

How can HR experiment with AI in low-risk, high-value ways?

Here are practical, manageable steps for HR teams to begin AI experimentation without compromising data security or employee trust.

1. Start with controlled pilot projects

Begin with small AI projects in low-risk areas, such as internal data analysis or automated responses for frequently asked HR questions. Jess Larsen’s experience as VP of People at open banking fintech TrueLayer provides a valuable blueprint here. When her team began using tools like Gemini and ChatGPT they limited the types of data entered and set specific ground rules around data confidentiality. By playing with controlled exercises like writing scripts or analysing survey results TrueLayer’s team could test the waters of AI without jeopardising sensitive information. This type of thoughtful experimentation allows HR teams to evaluate AI’s potential while mitigating risks, something any organisation can replicate. Such controlled projects don’t just provide useful learnings; they also build confidence across the team, helping HR staff become more comfortable and creative with AI applications. As Larsen says, even minor experiments can spark “light bulb moments,” demonstrating the ways AI can amplify everyday tasks like adjusting tone for different audiences or parsing survey results for insights. 

2. Plot tasks by value and risk

As The People Space’s editorial director Siân Harrington suggests HR teams can plot tasks along a value-risk axis to find ideal starting points. Tasks with high value and lower risk – like data visualisation and CV parsing – are excellent candidates. This way HR teams can focus on areas where AI can provide immediate value without exposing the organisation to unnecessary risk. Her advice? Don’t wait. Start small, experiment in a safe space and let the outcomes guide your next steps.

3. Empower frontline and ‘accidental’ managers

One of the most promising areas for AI in HR lies in the support of frontline managers. In fact, as Devoy noted, the UK has approximately 2.4 million “accidental managers” – people promoted to managerial positions without formal training. This group is typically thrust into leadership roles without sufficient support, often resulting in lower employee engagement and higher turnover. AI, however, can step in to fill gaps in knowledge and skills, providing managers with valuable assistive tools that streamline complex processes, give performance insights, and even act as on-demand management coaches.

Take Devoy’s own experience: using Microsoft’s Copilot feature he found he could prepare board papers in less than an hour, a task that would have taken most of a workday previously. With Copilot handling 80% of the drafting work Devoy could focus on adding strategic insights, effectively transforming the way he manages his workload. Similarly, frontline managers can leverage AI to assist with everything from project planning to employee evaluations, enabling them to devote more time to high-value tasks, like team building and employee engagement.

Larsen agrees that AI offers a major advantage for managers, handling much of the groundwork typically covered in “101 management training”. By providing quick, scenario-based advice AI tools can help managers resolve day-to-day challenges with guidance that’s 70-80% as effective as formal training. This ‘point of need’ support reduces hesitation to seek help, giving managers a ‘secret aid’ for navigating issues they may feel unsure about, especially valuable for those lacking formal training or confidence. AI, therefore, can be an immediate level-up tool, equipping managers to perform more effectively in real-time.

How can AI improve fairness in hiring?

In recruitment AI can be especially impactful, helping companies mitigate bias and improve decision-making processes. Danny McGrath, senior manager customer success at recruiting platform Greenhouse, highlights an example that any HR leader can relate to. Using an interview intelligence tool called BrightHire, his team at Greenhouse is able to record, transcribe and summarise interviews. This provides hiring panels with objective data, ensuring that decisions aren’t overly influenced by any single voice. “It's made a really, really big impact to both the speed of the hiring process as well as the quality of hire that we're seeing come out of the process at the other side,” he says. 

Beyond BrightHire Greenhouse also anonymises candidate information in CVs, removing identifiers like name, gender and educational institutions to ensure a fairer review process. This deliberate approach allows AI to enhance both the speed and quality of hiring, aligning perfectly with HR’s commitment to equitable practices.

To those in HR still hesitant to bring AI into recruitment McGrath’s example shows a clear, practical application that not only improves hiring efficiency but also supports diversity and inclusion goals. By ensuring that decisions are based on candidate qualifications alone HR can begin to reshape hiring in ways that prioritise skill and potential over subconscious bias.

What safeguards should HR establish for AI use?

Setting clear guidelines is essential before implementing any AI tools in HR. Start by establishing basic data security measures and consider using AI tools that are integrated within secure, licensed environments. At his own organisation Devoy, for instance, uses Copilot within Microsoft’s secure environment, ensuring that data is safely contained.

Key safeguards include:

• Setting boundaries: Limit data input to non-sensitive information unless securely contained
• Defining acceptable use: Clearly outline what data and tasks are appropriate for AI
• Collaborating with Legal and IT: Work cross-functionally with compliance, legal and IT teams to set up secure, compliant workflows for any AI experimentation.

What are the key takeaways for HR leaders experimenting with AI?

1. Think big, start small

   As Paul Devoy says: “Don’t get left at the harbour.” Start small, low-risk experiments now to experience AI’s benefits without waiting for full-scale implementation.

2.  Focus on high-value, low-risk tasks

   Begin with tasks where AI can quickly add value without major risk. Use a value-risk matrix to identify opportunities

3. Upskill HR teams on AI basics

   AI is a powerful tool but it requires informed users. Train HR teams on AI basics to ensure they can manage its risks and maximise its potential

4. Prioritise fairness and compliance

   Implement AI in ways that enhance equity and meet regulatory standards, particularly in recruitment and data use.

Embrace AI thoughtfully for lasting impact

The ship has indeed set sail and HR leaders who delay engagement with AI risk being left behind.  As Harrington warns, most of HR is not on the ship at the moment so starting small and learning from early experiments is crucial if it is to catch up with, and keep pace in, this new era of workplace technology.

AI doesn’t have to be intimidating or risky. With purpose-driven experimentation and robust safeguards HR leaders can make AI a constructive part of their operations, boosting productivity, supporting managers and creating fairer hiring processes – thereby ensuring AI is a tool for people-centred innovation, augmenting and not replacing workers.