Future of Sourcing and AI: The Tipping Point Is Happening NOW

Introduction

After more than 20 years in recruitment — including 10 years entirely dedicated to sourcing, evolving the discipline, exploring countless tools, studying market shifts, and anticipating what’s next — I’m picking up the pen again today because we’ve reached a major inflection point.What’s happening right now will determine the very future of our profession.

LinkedIn won the Game, there is no more game.

Back in 2018, during a talk at Sourcing Summit UK, I introduced the “pedestrian street theory.” It compared recruitment to a pedestrian street where you have to attract passers-by to hire them.I highlighted something fundamental: with its massive market penetration, LinkedIn had removed the anonymity of those passers-by.

For the first time, it became truly possible to source a huge portion of the professional population — as long as you knew how to use the tools.

What makes LinkedIn both essential and irreplaceable is that 1.1 billion people have manually entered their data into the platform, directly linking themselves to their employers and creating structured profiles.Job boards, on the other hand, merely collected CVs — unstructured data. That structural difference completely changes the game and enables an unprecedented level of precision in searches.

In Europe and elsewhere, the competition is over: Viadeo is dead, Xing has been relegated to the amateur league.LinkedIn won. The monopoly is total.

The AI Tsunami Is Rising

ChatGPT is only three years old. And in just three years, the leaps in artificial intelligence have completely transformed how I work, think, and envision the future of our field.The evolution is absolutely exponential — and believing that we’ve reached any kind of technological maturity would be naive.This is just the beginning.

I’ve seen AI technologies — in voice or outreach message generation — that leave a large portion of the recruiter population far, far behind.And these are still their worst versions.

We’re in November 2025. In a year, I believe everything will have changed — and no one will doubt that an AI suite can replace a significant share of a sourcer’s or recruiter’s skills.

All of this evolution hinges on one thing: LinkedIn data availability — or lack thereof.

LinkedIn’s Business Strategy

LinkedIn doesn’t give access to its data — it’s their war chest. No one has full API access to query the database, and in a way, that’s understandable.

LinkedIn’s commercial strategy is built on strong, steady, continuous price increases, with no intention of slowing down. They have:

• Made InMails indispensable for large companies

• Removed access to users’ own network data (try downloading your network — you’ll get very little)

• Made extracting data from the platform nearly impossible

• Created the single hub of structured, searchable professional data

Simple exercise: look at the last 30 people you hired. Now that you have their names — how many didn’t have a LinkedIn profile? (Again, finding them is another matter.)

This year, LinkedIn’s clear focus is on its “hiring assistant.” Technically speaking, though, LinkedIn still has massive shortcomings in deploying any serious AI — despite having the best dataset in the world.

For those who remember the BCG matrix: LinkedIn is in its Cash Cow phase, and they clearly plan to milk it.

The “Magic AI Tools”

Since last spring, tools claiming to do external sourcing and replace sourcers have been popping up everywhere — in the U.S., Europe, even Switzerland.

Every single tool I’ve tested that claims to perform sourcing clearly uses LinkedIn data.Some even display, in their demo videos, the number of connections and followers of profiles. Tell me that’s not LinkedIn data.

Without being a GDPR expert, it’s easy to see these tools shouldn’t have access to that data — it belongs to LinkedIn. Every time I’ve asked these companies what their legal basis was for holding such data on their servers or via data brokers, I got nothing but awkward smiles or evasive answers.

The Tipping Point

The tools I’ve tested so far range from “nonsense” to “not great,” but I’m starting to see promising things.Knowing what I know about AI advances and LinkedIn’s dataset, it would be foolish to think that — with such technology, millions (or billions) in funding, and skilled teams — we wouldn’t soon be able to replace a huge part of the sourcing function.

I spent months earlier this year deeply exploring LinkedIn Recruiter’s search features, which led me to create the most advanced training available on the subject. I fully understand how easy it would be to outperform a large number of recruiters who use the tool at a very basic level.

Frankly, even LinkedIn’s internal AI is an embarrassment to anyone who understands sourcing and how that database actually works.

So to me, whether these tools work now isn’t the point — it’s only a matter of time before they do.

A Call to the Community

I have no desire to risk a defamation lawsuit. Many of these tools loudly claim they’re fully compliant with all legal requirements. I’m just waiting for proof.

So I’m calling on GDPR experts, data management specialists, and DPOs to clarify whether these tools are legally legitimate. None of the ones I contacted were willing to reveal their data sources or clearly explain their legal grounds.

I’ve consciously chosen not to share a list of tools here — even though I’ve spent a lot of time compiling one and tracing potential providers and data brokers.

Please tag the relevant people. I’d be happy to create a private group. I’ve already reported as much as I could through my LinkedIn contact — but Guillaume Alexandre, in his quiet Swiss watchmaking valley, isn’t exactly their top priority. (When I was part of a GAFAM internal group that liaised directly with LinkedIn, I did feel a bit more listened to, lol.)

Don’t hesitate to involve lawyers, LinkedIn reps, or regulators — anyone who can help set the record straight.

A Computer Costs Less Than a Salary

Even if these systems aren’t perfect — recruiters aren’t either.

Whatever the cost of an AI system that works even halfway, it will still be negligible compared to a human salary.There’s far too little expertise in our profession for decision-makers not to see the appeal of simply replacing sourcers and recruiters with AI.

Given the poor response rates, search quality, and messaging seen from some recruiters, it’s fair to say that even a poorly configured computer wouldn’t necessarily do worse.Many recruiters have fallen for an illusion of mastery — the interface looks easy, but it hides deeply flawed logic designed by people who don’t understand recruitment.

Example

Without going too technical: LinkedIn’s AI pushes you to use the “industry” filter. That’s absurd — it’s not mapped to the company’s sector but self-declared by the user, who usually never updates it when changing jobs. My calculations show 30% to 60% of missing or incorrect results from that single filter.By tweaking just two autofilled filters, we’ve seen 2.7x more profiles.If that intrigues you — that’s a tiny glimpse of why advanced sourcing training still matters.

It’s Time to Draw the Line

We need legal clarity on whether these systems have the right to possess and use LinkedIn data.

If their legal basis is confirmed valid — I’m sorry to say the average recruiter or sourcer’s days are numbered. Some experts will still go beyond what the machine can do, but for how long? It won’t be an industry anymore — just isolated individuals.

To those who think “machines can’t replace human contact” — think again.Once the system works, all it will take is a freemium model where HR just feeds in the job description, does an intake meeting, and the machine handles all sourcing automatically.“You only pay if you hire.”

Of course, I’m not talking about replacing real interviews or structured, professional recruitment processes. I’m talking about the transformation of people into candidates.

And to the confident ones saying, “People don’t want to be contacted by machines”:I can assure you the upcoming technologies are more professional and empathetic than most recruiters I’ve met. They work 24/7 and add real value for candidates.If the experience is cheaper, faster, more efficient, and satisfying for both sides — remember, recruiters are just the interface connecting a CV with an offer.

The Trap for Recruitment Agencies

Given LinkedIn’s pricing strategy, it’s easy to see why many turn to alternatives — they’re cheaper and might even work better.But what buyers of these solutions fail (or refuse) to see is that soon it’ll be enough to buy a license — and the whole sourcing process will be automated.I don’t think that’s years away — it’s probably months.

Some tools even proudly claim:

“Sourcing sucks — let AI do it for you.”

The temptation is strong for agencies wanting to cut costs. But by adopting these systems, they’re introducing a technology that will soon allow their clients to bypass them entirely at a fraction of the price.

Don’t think you’re irreplaceable. These systems are being built to automate your work from end to end.

If LinkedIn or Regulators Draw a Clear Line

If they declare these practices illegal, then mastering LinkedIn Recruiter will continue to hold value — and LinkedIn will be free to keep raising prices and tightening control over recruitment.

A Bit of Foresight

After many conversations (thank you, you know who you are), I can’t believe LinkedIn is unaware of what’s happening.They’re not that incompetent — nor that careless about their monopoly.

The only realistic explanation — cynical as it is — is that LinkedIn knows these tools are scraping its data, knows its own AI efforts are underwhelming, and is simply waiting.

My Hypothesis

They’re waiting for one player to crack the code — to build something that genuinely replaces a large part of sourcing and recruiting.Once that happens, LinkedIn will buy that company, shut everyone else down, and take legal action to eliminate competitors.They’ll emerge with complete monopoly and total control over the market.

There are even rumors that LinkedIn is pushing three-year contracts, knowing that once you’ve signed, you can’t get out.That could mean LinkedIn doesn’t even expect humans to still be doing this job three years from now.But that’s just a rumor — I haven’t confirmed it myself.

Now Means Now

It’s time to hold LinkedIn accountable — to demand a clear stance on these tools.

Legal or not?Regulators must also weigh in: legal or not?

Yes, that might mean supporting an already dominant monopoly — but at least it could delay the obsolescence of our profession.

The question is no longer if AI will transform sourcing — but when, how, and above all, who will control the data powering that revolution.The tipping point is happening now.

And no — I’m not announcing the end of sourcing or sourcers. There will always be room for specialists, for bespoke work, for niche expertise.But the disappearance of the bulk of sourcing as a human-led function?Yes.