Author: Piotr Lawrynowicz

A Brief Recap of Earth-Shattering News From the Past 20 Years:

1. The 1980s – the first home computers arrived (Sinclair, Commodore, Atari, Amiga, and friends). Scientists warned: “Young people will get used to an unreal world and society will lose its mind.”
   
   Indeed, society did lose its mind – though not because of computers, but thanks to bad sitcoms.
2. Also the 1980s – The Terminator hit the cinemas, kick-starting a fashion for Judgment Day stories – self-learning systems hell-bent on wiping out humanity. It was, in many ways, a remarkable production. Only one small detail: back then people knew about IT about as much as a dog staring at the moon.
   
   They actually knew more about outer space – yet to this day no one has delivered on any grand plan to colonise the Moon or Mars.
3. The 1990s – the apocalypse was cancelled. IT moved on. We got the productivity revolution and miniaturisation. Computers became ever smaller, ever faster. The all-knowing voices predicted the next catastrophe: “People will have brain chips implanted, and everyone will be controlled by a supercomputer called THE BEAST.”
   
   As far as I know, I don’t have a chip in my brain – I’ve been in IT since 1996 and have yet to hear of a supercomputer running our thoughts. I did, however, once see a lift system in Warsaw allegedly based on something resembling a neural network. To this day I’m not sure what that means – perhaps someone mixed things up and, instead of installing chips into brains, they started putting them in elevators.
4. The 21st century – after the Y2K bug and other minor episodes (not worth the belly laugh), the noise died down. Computers evolved, systems became friendlier, all doomsday scenarios were called off.
   
   Then about 10 years ago, Big Data showed up, followed by tools to analyse it – a short hop to AI (and yes, the first AI films had already been made).
5. The past 7–8 years – theories about AI eliminating humans gained traction, because computing power kept growing and systems kept analysing – only now with much greater horsepower. It’s a topic that’s loud, media-friendly, anchored in pop culture. Some see AI as pure evil; others want to bet the farm on it because humans are “less efficient”.
   
   This article is for those who choose The Reasonable Option.

The Great Misunderstanding of 2025

A few days ago I was in a meeting with the CFO of one of Poland’s largest companies.
“Piotr,” he said, “why do I need HR outsourcing when ChatGPT can already write job descriptions and recruitment algorithms screen CVs better than humans?”
I smiled. That’s like asking: “Why do I need a mechanic if I’ve got GPS?”
GPS will show you the way, but it won’t fix your engine when you break down in the middle of a potato field.

The Truth About AI in HR: A Tool, Not an Employee

AI isn’t a superhero that’s going to send your HR department into early retirement. It’s more like a super-tool that still needs someone who knows how to operate it.

Imagine you’re handed the most advanced surgical equipment in the world. Would you immediately perform open-heart surgery? Of course not (unless we’re talking about an autopsy, in which case, good luck). You’d need a surgeon who:

• knows when to use it,
• understands its limitations,
• can interpret the results,
• takes responsibility for the decisions.

Exactly the same applies to AI in HR.

The New Reality: AI Creates More Work, Not Less

1. Someone Has to “Train” the Algorithms
   Your AI recruitment system is like a puppy – at first it makes more mess than it’s worth. It needs:
   • proper data (and how many companies have clean, structured HR data?),
   • constant fine-tuning (algorithm discriminates against women? Oops.),
   • learning material,
   • result verification (did AI really find the best candidate?).
   At SmartPeople (yes, I’m quite happy to say it’s the company I work for and develop), we see this daily. Example? Here you go: a client implements AI to analyse performance reviews. After a month they call:
   “Piotr, the system shows our best employee is… the cleaning lady. What do we do??”
   Answer: you need specialists who know how to train AI.
2. Compliance in the AI Era Is a New Level of Hell
   Remember GDPR? That was child’s play compared to what’s coming.
   The EU AI Act (already in force!) requires companies to:
   • document every algorithm used in HR,
   • ensure the “explainability” of AI decisions,
   • audit bias in recruitment systems,
   • obtain employee consent for AI monitoring.
   How many of your internal HR people know these regulations – and how many have time to keep up with them?
3. Human-in-the-Loop Isn’t Optional, It’s Essential (!!!!!!!)
   AI can process CVs in seconds. But what about a candidate who:
   • has a five-year gap in their CV (depression? caring for a parent or child?),
   • changes industry every two years (unstable or adaptable?),
   • has an education unrelated to the position (outsider or diamond in the rough?).
   AI sees patterns. Humans see humans – and feel them. Has everyone in HR forgotten that it’s not just “Resources”, it’s first and foremost “Human”? In our projects (SmartPeople – yes, let’s say it loud), we use a hybrid approach: AI screens, humans decide. Recruitment effectiveness jumps by 40% – but you still need experienced recruiters who know when to ignore the algorithm.

Why HR Outsourcing Will Grow Thanks to AI

The Digital Photography Analogy

Remember when digital cameras arrived? Everyone thought: “That’s it for photographers! Anyone can take a picture now!”
What happened? There are more photographers than ever.
Why? Because easy tools increased demand for professional quality. Everyone takes pictures – professionals do it well.

The Same Thing Is Happening with HR and AI

Before AI: companies had basic HR processes, did things “somehow”.
After AI: companies see they can do it better, faster, with data. They want that level – but lack the competence, imagination, and perspective.

Where Exactly HR Outsourcing Gains from AI

1. AI Implementation & Management
   Our financial-sector client was rolling out AI for employee experience. A project planned for three months took fourteen. Why? Because no one internally knew:
   • how to prepare the data,
   • how to define success metrics,
   • how to integrate with existing systems.
   Result: they now outsource entire AI projects to us. Fourteen days to go-live, ROI guaranteed.
2. Data Analysis & Insights
   AI generates mountains of data. Does your HR Business Partner have time to analyse 47 dashboards a day?
   Our teams deliver actionable insights, not raw numbers, e.g.:
   • “Peter has a 73% probability of leaving in Q2 – here are three retention actions.”
   • “Your recruitment process is 23% slower than the benchmark – here are the bottlenecks.”
3. Continuous Learning & Adaptation
   AI doesn’t self-evolve. It needs people who:
   • track new regulations,
   • test new tools,
   • benchmark against the market,
   • adapt processes.
   Problem: your internal HR doesn’t have time for R&D – they’re busy with daily operations.
   Solution: you outsource to teams who live on the bleeding edge.

ROI That CFOs Will Love

One of our clients (manufacturing company, 800 employees) compared costs:

The Future Is Already Here

Don’t ask: “Will AI replace HR outsourcing?”
Ask: “How quickly can I implement AI-powered HR outsourcing so my competitors don’t leave me in the dust?”

Because while you’re still wondering whether to recruit in-house AI specialists, your competitors are already:

• automating 67% of recruitment processes,
• predicting employee churn with 89% accuracy,
• cutting time-to-hire by 45%.

Bottom Line

AI will not replace HR outsourcing. AI makes HR outsourcing absolutely essential.
Why? Because AI isn’t “set and forget”. It’s “set, tune, monitor, adapt, improve” – forever.
And who has the time and competence for that? Not your internal HR, tied up in day-to-day firefighting.
Specialists who live and breathe AI. Who test, learn, adapt – and deliver results, not experiments.

Want to see how AI-powered HR outsourcing could transform your company?
14 days to first results, 7 months to payback, full KPI transparency.
DM or email: piotr.lawrynowicz@smartpeople.com.pl

PS: The best CFOs in Europe already get this. The only question is: will you be in that group – or in the one still “thinking about AI in HR”? (Both answers are fine, but only one will save you the time and money your company actually exists for.)