“AI says my role pays X.” Calibrating compensation in the era of noisy benchmarks
📝 From the editor
👀 AI is now in the room when we talk about pay.
Candidates and employees are quoting salary ranges from ChatGPT-style tools, hoping for more transparency and fair compensation.
In theory, that should help. In practice, many figures are scraped from unreliable sources (like anonymous self-reported data), detached from regional markets, and misaligned with how your company defines scope and leveling.
This isn’t an edge case. Generative AI is now a mainstay for knowledge workers, and forward-looking companies encourage thoughtful adoption (see our “Just use AI” isn’t a strategy edition for practical guidance). That’s positive. But AI-generated numbers now crop up earlier and more often in the hiring process.
The challenge for People leaders: how do you keep pay talks fair and consistent when unverified data can drown out your internal logic?
Regulatory momentum raises the stakes. EU countries must implement the EU Pay Transparency Directive into national law by June 7, 2026. From the first reporting cycles after these national rules take effect, large employers will be required to publish gender pay gap data, with smaller companies phased in later. If you hire in the EU, you’ll soon be held to higher standards for clear salary bands, documentation, and explainable decisions.
Fairness goes beyond compliance. A recent study (🎩 hat tip to Jillian Climie — see The People Pick below 😉) found that large language models recommended dramatically lower pay for women than men, even when given identical roles and locations.
The point isn’t to “bash” AI, but to remember that its outputs mirror its training data. Fair outcomes start with fair inputs. That means reliable market data, transparent pay structures, and clear career frameworks. Not numbers a chatbot pulled without context.
HR operates at the crossroads of performance, progression, and pay. What works is to anchor compensation to well-governed bands, link pay to level expectations and evidence from well-structured reviews, and equip managers to explain the “why” openly and confidently.
In this issue, we unpack that playbook: how to calibrate pay in an AI-noisy world, and why a unified approach to hiring, reviews, and promotions is the best defense against confusion, bias, and drift.
Scroll for practical moves you can implement now, plus clear visuals to make them easier to put into action.
P.S. You’ll hear from me every other week. Expect research-backed insights and fresh perspectives. You’re also shaping this journey, so reach out anytime with topics you’d like us to cover.
.
🔍 TheLeapsome lens
Comp in an AI world: 5 moves to keep pay fair
1️⃣ Define “the market” in writing
Document your pay data sources, which market percentiles you target, your location policy, and how often you update benchmarks. Make these decisions explicit and embed them in your salary processes.
Every hiring, review, and promotion panel should reference a visible, up-to-date career framework. That consistency builds trust internally and externally.
3️⃣ Support HR and managers to thoughtfully navigate “AI said X”
Encourage HR colleagues and all managers to approach these discussions with curiosity and fairness by asking: “Where does this number come from? Which data inputs, which market, and what timeframe?”
Then, guide the conversation back to transparent salary bands and career frameworks that reflect both the employee’s role and development path. This approach supports trust and centers the employee’s experience.
Routinely analyze pay data to spot out-of-band exceptions, gaps by org level, or disparities by location. Proactively close these gaps, not just after the fact.
5️⃣ Link reviews to compensation (with guardrails)
Integrate performance and development data into compensation reviews. Stay transparent, guard against pay compression (levels paid too similarly) or bias, and always make decisions explainable.
Pay doesn’t happen in a vacuum. With Leapsome, compensation, reviews, and career frameworks work hand in hand — together with the rest of your People processes — in one platform.
This makes every decision fair, explainable, and trusted.
In each issue, we highlight powerful perspectives shaping how we work.
📺 All Things Work by SHRM: “The truth about pay transparency: challenges, risks, and rewards
The episode looks at how transparency is reshaping workplaces — and why success requires more than compliance. Expect guidance on defining bands and criteria, preparing managers for tough questions, and communicating proactively to address gaps and build trust.
Featuring guests Nanci Hibschman (Managing Principal, SullivanCotter and C3 Nonprofit Consulting Group) and Amanda Wethington (Principal, C3 Nonprofit Consulting Group).
📝 Harvard Business Review: “New research debunks a common criticism of pay transparency”
Mary Ellen Carter, Lisa LaViers, Jason Sandvik, and Da Xu analyzed data from 1,300+ publicly traded firms and found that well-designed transparency can increase compensation satisfaction — especially when paired with clear structures and communication. Useful context when AI-generated ranges enter pay talks.
Today, Jillian Climie — gender equity expert and Co-Founder of The Thoughtful Co. — shares a crucial insight with us:
Research has shown that large language models (including tools like ChatGPT) have recommended significantly lower salaries for women than men, based on otherwise identical prompts specifying position, location, and year. In one study, the suggested pay differed by US$120K.
This is a clear, worrisome reminder that these systems currently reflect biases in their training data.
AI can be a powerful tool, but only if HR leaders understand its limitations and question the outputs instead of accepting them at face value.
📣 Events worth your time
🗓 Coming up: Meet us at Zukunft Personal 2025
Sept. 9-11 | Cologne, Germany
We’ll be at Europe’s leading HR event, reimagining the future of work with thousands of People experts. Swing by our booth, catch a live demo, or chat with us about your HR strategy.
▶️ On demand: “Defending People-First HR Strategies in 2025”
Watch our deep dive into Leapsome’s 2025 HR Insights Report. Learn how HR leaders are navigating economic, cultural, and technological pressures without deprioritizing employees. You’ll discover…
That 92% of HR leaders report pushback, usually due to a lack of clear “why.” Learn how purposeful storytelling shifts buy-in.
How rigid RTO mandates risk trust, and why transparent, collaborative approaches win. [For a deeper dive, read our RTO edition, Presence ≠ Performance.]
The growing strategic leadership role of HR in workforce and AI planning.
We are in a day and time where we must be the people who step forward first. Intentionally. We have to be, you don’t need permission. You have permission after this webinar (...) because if you don’t, you’re letting your company fail.
— Steve Browne, Chief People Officer, LaRosa’s, Inc.
![POPs header 600x200 [email]](https://leapsome-5468055.hs-sites.com/hs-fs/hubfs/POPs%20header%20600x200%20%5Bemail%5D.png?upscale=true&width=1120&upscale=true&name=POPs%20header%20600x200%20%5Bemail%5D.png)
