Finding candidates isn't the hard part. Certainty is.
AI made finding people faster for most US recruiters, and for far fewer British ones. On both sides of the Atlantic, recruiters still can't tell who can do the job.

01.1
01.1
We polled 1,428 US recruiters and 483 UK recruiters on what AI changed for them in the last 12 months. In the US, one answer came back louder than the rest. Finding people has got faster. Nearly three-quarters of US recruiters, 72%, say AI increased the speed of finding candidates over the last 12 months.
In the UK, recruiters feel less of AI's impact. 58% (vs 72% in the US) say AI made finding candidates faster, four in ten (41%) say it gave them no speed gain at all, and more than one in five (22%) say it slowed them down. Only 13% of US recruiters say AI slowed them down.
But ask recruiters in either country whether they're any surer about those people they find, and it tells a different story. In the US, 59% say AI increased their confidence that they'd found the right person, but 18%, almost one in five, told us it’s made their confidence fall. In the UK, fewer than half (48%) say AI made them surer, and nearly a quarter (24%) say it made them less sure.
Of the US recruiters who say AI made finding candidates faster, around a third report no gain in certainty at all. In the UK, it's 45%. Put speed and certainty together, and 48% of US recruiters say AI made them both faster and surer. In the UK, it's a third (32%).
AI increased the speed of finding candidates
AI decreased the speed of finding candidates
AI increased their confidence
AI decreased their confidence
AI made them both faster and surer
AI increased the speed of finding candidates
AI decreased the speed of finding candidates
AI increased their confidence
AI decreased their confidence
AI made them both faster and surer
0 %
in the US say they have no clear picture of what “good” looks like in an AI role
0 %
in the UK say they have no clear picture of what “good” looks like in an AI role
So recruiters can now source faster, in the US at least. But are they any surer about who they find?
In this report, we take a pulse check on recruiting in 2026 in the US and the UK. We will dig into this split, why it has occurred, why it looks different on each side of the Atlantic, and what can be done about it.
When it comes to sourcing, nearly every tool competes on speed: more profiles, faster searches, automated outreach at scale. Many problems with sourcing that existed for years now seem largely solved.
However, the problem recruiters name as their hardest, certainty in the candidates they find, hasn't moved at all. It ranked first in the US in 2025, it ranks first again in 2026, and it ranks first in the UK too.
Let’s dig into why.
02.1
02.1
of the 1,428 US and 483 UK talent professionals surveyed name verifying whether resume skills are real as a top-three sourcing challenge, the single most cited challenge in both countries.
believe half or more of the candidate profiles and resumes they see are written by AI. 53% of US and 44% of UK recruiters believe all or most are.
say that when a sourced candidate's profile claimed AI skills, those skills proved weaker than the profile suggested, at least sometimes. 54% in the US and 45% in the UK say it happens often or almost always.
believe most or all of the candidate data in their sourcing tools has been independently tested rather than self-reported.
02.2
02.2
say AI increased the speed of finding candidates in the last 12 months. 22% of UK recruiters say it slowed them down, against 13% in the US.
say AI increased their confidence that they had found the right person.
say AI decreased their confidence that they had found the right person.
say AI made them both faster at finding candidates and surer they had found the right one.
say the hardest part of judging a candidate for a new AI-specific role is that there is no track record to read. It's the top answer in the US and third in the UK.
say every profile for these roles reads the same. In the UK, it's the joint top answer.
0 %
in the US say they have no clear picture of what good looks like in an AI role.
0 %
in the UK say they have no clear picture of what good looks like in an AI role.
02.3
02.3
review 26 or more candidate profiles to find one worth contacting.
hear back from fewer than a quarter of the candidates they contact directly.
spend an hour or more of their own time on a single candidate before they trust that person can do the job. 29% in the US and 30% in the UK spend two hours or more.
02.4
02.4
would be willing to run some form of objective check on a sourced candidate before interviewing them.
would use a short job simulation. 51% in the US and 43% in the UK would use a 10-minute skills test.
would ask a candidate directly and take their word that they can do the job.
would go straight to interview with no check at all.
02.5
02.5
1,428 US talent professionals, fielded in August 2026, and 483 UK talent professionals, fielded September 10 to October 7, 2026. Both samples used the same questionnaire, exclusions, and weighting: 55% talent acquisition, 25% external recruiter, 20% HR leader. US analysis based on 1,428 drawn from 2,205 completes.
0 %
talent acquisition
0 %
external recruiter
0 %
HR leader, in both countries
We asked recruiters in the US and the UK to tell us where sourcing breaks, and five findings came back. Most hold on both sides of the Atlantic. Where they don't, the difference is telling. Let’s break them down:
03.1
03.1
Just over half of US recruiters, 53%, believe all or most of the resumes and profiles they see are written by AI. In the UK, 44% say the same. Widen it to half or more, and the two countries are level, at 73% in the US and 75% in the UK. Three in four British recruiters think AI wrote at least half of the CVs they see.
0 %
of recruiters in the US believe half or more of the resumes and profiles they see are written by AI
0 %
of recruiters in the UK believe half or more of the resumes and profiles they see are written by AI
This stat, while eye-opening, isn’t entirely damning. A well-written resume has never really been proof of anything, and helping people present themselves clearly is a genuinely helpful thing for a tool to do. The problem is what happens to the claims inside.
Verifying that the skills on a resume are real was named by 57% of US recruiters as a top-three challenge, ahead of “a shortage of skilled candidates” and “culture alignment”. It ranked first when we asked in 2025 as well. UK recruiters put it first too, at 49%, ahead of a shortage of skilled candidates at 43%.
British recruiters are also harder on what their current sources give them. They're more likely than US recruiters to name poor-quality candidates from existing sources (25% against 17%) and ineffective sourcing tools (15% against 10%) as a top challenge.
AI has only compounded the problem. 54% of US recruiters say that when a sourced candidate's profile claimed “AI skills”, those skills turned out weaker than the profile suggested, “often” or “almost always”. Nearly nine in ten say it happens at least “sometimes” in both countries: 90% in the US and 87% in the UK.
The claim doesn't fail at sourcing so much as survive it. We asked hiring leaders earlier this year on the state of hiring for AI fluency, and 59% said they'd made a bad AI hire in the previous 12 months: i.e., someone who sounded fluent in the interview and couldn't apply it on the job.
This is the pain point our own customer conversations surface most often, and the survey puts a hard statement on it: you can't trust what you see. British recruiters have taken that further. Only 15% would take a candidate's word that they can do the job, against 23% of US recruiters.
0 %
of recruiters said they'd made a bad AI hire in the previous 12 months
The signal recruiters have always sourced on, the written profile, is now produced by the same category of technology that made searching those profiles fast. Recruiters got a better search engine and a worse index in the same 12 months.
The findings also reframe what "faster" delivered. If AI doubles the number of profiles you can assess in an hour, and a rising share of those profiles overstate what the person can do, the speed gain is mitigated by a lower-quality pool.
Want to get more insight? We dived deeper into the trustworthiness of the AI-written resume
03.2
03.2
Beyond the resume's validity, another major problem seems to have gotten worse in 2026. In our survey, US recruiters naming a shortage of skilled candidates rose 9 points from 2025 to 51%. In the UK, 43% name it, second only to verifying skills, the same order as in the US.
0 %
US recruiters naming a shortage of skilled candidates rose 9 points from 2025 to 51%
AI made searching faster for 72% of US recruiters and 58% of UK recruiters. Most candidate pools can be swum through in one afternoon. Back when a search took days, scarcity was hidden inside the work of looking. You didn't know whether the right people weren't there or whether you simply hadn't gotten to them yet. Now that the search can be done rapidly, that deep pool is suddenly starting to look pretty shallow.
Nothing has necessarily changed about how many qualified people exist. What changed is that recruiters can now see how few there are.
Pool size and specialist role scarcity aren't independent problems. Faster search converts an abstract shortage into a concrete one, and it does it without adding a single qualified candidate to the market.
The UK adds a twist. Among recruiters working mainly on hard-to-fill specialist roles, 60% in the UK say AI made them faster, against 78% in the US. The searches where speed matters most are the ones where British recruiters say AI has helped least, so more of their scarcity is still hidden inside slow searching.
03.3
03.3
Specialist role scarcity gets sharper still when the role itself is new.
More than a quarter of US recruiters (28%), and one in five UK recruiters (20%), say the people who can do the work simply aren't visible on the platforms they use. That's not a ranking problem or a filtering problem. It's an inventory problem, and it's the clearest statement in this study that reach and relevance are different things. You can have access to hundreds of millions of profiles and still not have access to the fifty people who can do a specific new job.
For the candidates who are visible, the difficulty isn't finding them. It's having nothing to judge them against.
In the US, almost half (46%) say there's no track record to read, which is what you'd expect for roles that are barely two years old. 38% say every profile for these roles reads the same, which is what happens when a field is new enough that everyone lists the same four tools and the same three certifications, while 35% say they have no clear picture of what good even looks like.
British recruiters name the same three problems in a different order. Their joint top answers are that every profile reads the same (33%) and that there's no clear picture of what good looks like (33%). No track record comes third, at 31%. US recruiters are short of history to check. UK recruiters can't tell the candidates in front of them apart.
Their organizations think they've answered that question. 95% list AI fluency as a hiring requirement and 71% have formally defined what it means, so the definition exists somewhere in the business. It just isn't reaching the person running the search.
Read those together, and the picture starts to feel bleak in a very specific way. A third of recruiters, in both countries, are hiring for roles where they can't define the target, against profiles that are indistinguishable from one another, for candidates with no history to check, using tools that mostly can't see the right people anyway.
95% list AI fluency as a hiring requirement
95%
71% have formally defined what it means
71%
03.4
03.4
This is the finding we didn't expect.
63% of US recruiters believe most, if not all, of the candidate data in their sourcing tools has been independently tested rather than self-reported by the candidate. In the UK, 54% believe the same.
Among that confident US group, 65% also say claimed skills turned out weaker than advertised, often or almost always. Among recruiters who think half or less of their data is tested, only 36% report the same thing. That's a 29-point difference running in the opposite direction to the one you'd predict.
The pattern holds in the UK, though more weakly. Among confident UK recruiters, 51% say claimed skills turned out weaker often or almost always, against 39% of those who think half or less of their data is tested. That's a 12-point difference, pointing the same unexpected way.
41% of the whole US sample hold both positions at once. They believe their candidate data is verified, but they also report it failing them.
of US recruiters believe their candidate data is verified, but they also report it failing them
believe most, if not all, of the candidate data in their sourcing tools has been independently tested rather than self-reported by the candidate
of confident recruiters say claimed skills turned out weaker than advertised
of US recruiters believe their candidate data is verified, but they also report it failing them
believe most, if not all, of the candidate data in their sourcing tools has been independently tested rather than self-reported by the candidate
of confident recruiters say claimed skills turned out weaker than advertised
So in short, we have a paradox in that the recruiters who report being most confident in their candidate data are often the same ones who say that data let them down.
Two explanations for why this paradox is occurring:
If you believe the data in front of you has already been checked, you aren’t likely to check it again yourself. So, the claims go unexamined at sourcing, the candidate moves forward on the strength of the profile, and the mismatch only surfaces at interview, where it lands as a failure. Confidence at the top of the funnel produces disappointment at the bottom of it, and the disappointment is what gets remembered and reported.
Recruiters who trust their tools advance more candidates on tool-supplied evidence. More candidates advanced means more opportunities to watch a claim fall apart. Trust doesn't cause the failures so much as it increases how many you witness.
Both explanations end up in the same place. Trust in tool-supplied candidate data isn't being calibrated by evidence. Nothing in a typical sourcing workflow tells a recruiter which claims were checked, how, or by whom.
Can't trust what you see turns out to have a second layer: most recruiters can't see what they can and can't trust.
03.5
03.5
The manual grind is the pain point recruiters raise first in conversation, and this is what it's made of: high review volume, low reply rates, and no way to tell which of the profiles in front of them deserves the time.
Recruiters in both countries told us they’re still grinding through candidates just like before. Almost seven in ten, 69% in the US and in the UK, said that finding one person worth contacting still means working through 26 or more profiles.
This is what "faster" bought. AI may have compressed the time spent per profile, but it didn't reduce how many profiles stand between a recruiter and one credible candidate. A shorter time per unit, multiplied by the same number of units, is a smaller gain than the speed figure suggests, and it explains how 72% of US recruiters can report being faster while their productivity level feels largely unchanged. In the UK, fewer recruiters got the speed in the first place (58%), and the queue is the same length.
0 %
of US recruiters report being faster while their productivity level feels largely unchanged.
0 %
of UK recruiters report being faster while their productivity level feels largely unchanged.
The outreach doesn't reliably land either. A third of US recruiters, 33%, hear back from fewer than a quarter of the candidates they contact directly. In the UK, it's 40%. Work through 26 profiles, contact the best of them, and most of that effort goes unanswered.
04.1
04.1
Most recruiters in both countries tell us that they still verify skills themselves, by hand, one candidate at a time.
Two in three US recruiters (67%) spend an hour or more of their own time on a single candidate before they trust that person can do the job. In the UK, it's nearly three in four (73%). 29% of US recruiters and 30% of UK recruiters spend two hours or more.
That's the market's working answer to the verification problem right now. Recruiters have already concluded the question is worth answering. They're paying for the answer in their own hours because nothing in their stack answers it for them.
British recruiters pay more of that bill. AI gave fewer of them a speed gain to begin with, so they spend more of their own time per candidate, and fewer of them are willing to take a candidate's word for it (15%, against 23% in the US).
spend an hour or more on a single candidate before they trust that person can do the job
spend two hours or more.
spend an hour or more on a single candidate before they trust that person can do the job
spend two hours or more.
04.2
04.2
say they would be willing to run some form of objective check on a sourced candidate before interviewing them
would use a short job simulation
would use a 10-minute skills test
would run an ID or credential check
would take references
would ask the candidate directly and take their word for it
would ask the candidate directly and take their word for it
would go straight to an interview with no check at all
From this data, we see that resistance to proving a candidate's skill level isn't the obstacle in either country. If only 4% of US recruiters and 5% of UK recruiters would skip a check given a free one, the reason proof isn't being sought during sourcing is not that recruiters don't want it.
The clearest difference between the two countries is trust in the candidate's own account. Nearly a quarter of US recruiters (23%) would ask a candidate directly and take their word for it. In the UK, it's 15%, roughly one in seven.
04.3
04.3
Ask recruiters what makes sourcing hard, and you get the same four answers: the manual grind, the size of the pool, finding specialists, and trusting what they see.
and that's a real achievement. No recruiter in 2026 is short of profiles. Access to hundreds of millions of candidate records has gone from a differentiator to table stakes in about three years. Reach is no longer a competitive argument because it's nearly free.
AI took real time out of the work, and 72% of US recruiters saying “finding candidates got faster” is evidence of that. In the UK, the evidence is thinner, at 58%. And the compression happened per profile, not per search. 69% in both countries still work through 26 or more profiles to find one worth contacting, and 33% of US recruiters and 40% of UK recruiters still hear back from fewer than a quarter of the people they approach. The minutes per profile may have come down, but the queue hasn’t.
Faster search through the same inventory reaches the same people sooner. It doesn't add anyone. For 28% of US recruiters and 20% of UK recruiters, the specialists aren't on the platforms at all, which no amount of search speed will fix. What the technology delivered here was clarity about the problem, not a solution to it.
04.4
04.4
Identity, contact details, and employment history, for the most part. Sourcing platforms have invested heavily in confirming that a person is real, reachable, and worked where they say they worked. That work is useful, and it solved a real problem. But very few platforms can actually confirm that a person can do the thing their profile claims they can do, and fewer still will explain how their ranking arrived at the names it put in front of you.
So the category verifies the facts easiest to check against a database, and leaves the one that requires the candidate to demonstrate something. Meanwhile, the profile itself got less reliable. Over half of US recruiters now believe all or most profiles are written by AI, and three in four UK recruiters believe at least half are. The one problem the category didn't take on is the one that continues to get harder.
The two problems that more reach could solve were solved. But the two that require evidence rather than reach were not. Speed, scale, and coverage were the right answers to the questions the industry chose to ask, and none of them answer the question a recruiter is actually stuck on: can this person do the job?
04.5
04.5
Proof-first sourcing is the practice of treating evidence, rather than reach, as the measurement to limit a shortlist. It means criteria are agreed before candidates are scored, every name arrives with the reasoning attached, and a claim on a profile counts for less than a demonstration of the skill itself.
Recruiters got faster at finding people because the industry built for speed, and the UK shows how uneven even that gain has been. They didn't get surer on either side of the Atlantic, because certainty was never what the industry was building. The signal sourcing runs on is now written by the same technology that made searching it quick; the pool is visibly thinner than it used to look, and the one check that would settle the question is still being done by hand, an hour at a time, by the recruiter. In the UK, 73% spend that hour on a single candidate.
Reach has stopped being a competitive argument, because reach is close to free. 900 million profiles is not a meaningfully harder thing to offer than 100 million. The argument worth having now is whether you can defend the names on your shortlist: why these people, on what evidence, against which criteria, and what happens when a hiring manager asks.
That's a different way to build a sourcing tool, and according to our survey it's the one recruiters on both sides of the Atlantic are asking for. 92% in the US and 92% in the UK would be willing to run an objective check on a sourced candidate before interview.
05.1
05.1
TestGorilla is building the proof-first AI recruiting platform the industry needs, and Gorilla Sourcing treats reach and proof as separate problems.
On reach, you search across 900M+ candidate profiles, including 3M who are already skills-tested.
On proof, you set the search criteria, AI improves the search, and every name arrives with a written reason for its ranking. Then it's over to Gorilla Assess to test the shortlist. Nobody is assessed automatically, and no candidate sits an assessment before a recruiter chooses them, giving you total control over your shortlist.
How many recruiters struggle to verify candidate skills? 57% of the 1,428 US talent professionals and 49% of the 483 UK talent professionals in this study name verifying whether resume skills are real as one of their top three sourcing challenges. It was the single most cited challenge in both countries in 2026, and it also ranked first in the US in 2025.
What percentage of resumes are written by AI? We don't measure how many resumes are AI-written. We measure what recruiters believe: 73% of US recruiters and 75% of UK recruiters think half or more of the candidate profiles and resumes they see were written by AI. 53% in the US and 44% in the UK think all or most were.
Did AI make recruiting faster in 2026? Yes, for most recruiters, though far more in the US than the UK. 72% of US recruiters and 58% of UK recruiters say AI increased the speed of finding candidates over the previous 12 months, and 22% of UK recruiters say it slowed them down (13% in the US). Fewer say it increased their confidence that they had found the right person (59% in the US, 48% in the UK), and 18% of US and 24% of UK recruiters say their confidence fell.
How long does it take to verify one candidate's skills? 67% of US recruiters and 73% of UK recruiters spend an hour or more of their own time on a single candidate before they trust that person can do the job. 29% in the US and 30% in the UK spend two hours or more. This is self-reported time on one candidate, not an average across all hiring.
How many candidate profiles does a recruiter review to find one worth contacting? 69% of recruiters in both the US and the UK review 26 or more profiles to find one person worth reaching out to.
Would recruiters test a candidate before interviewing them? 92% in both countries say they would be willing to run some form of objective check on a sourced candidate before interviewing. In the US, a short job simulation and a 10-minute skills test tie at 51% each. In the UK, a short job simulation leads (46%), followed by a 10-minute skills test (43%). Only 4% of US and 5% of UK recruiters would go straight to interview. This measures willingness, not current practice.
Why is it hard to hire for AI roles? In the US, 46% of recruiters say there is no track record to read, 38% say every profile for these roles reads the same, and 35% say they have no clear picture of what good looks like. 28% say the people who can do the work aren't visible on the platforms they use. In the UK, the joint top answers are that every profile reads the same (33%) and there's no clear picture of what good looks like (33%), followed by no track record (31%). 20% say the right people aren't visible on their platforms.
Are UK recruiters more skeptical of AI than US recruiters? On speed and certainty, yes. 58% of UK recruiters say AI made finding candidates faster, against 72% in the US. Of those it did speed up, 45% in the UK saw no gain in confidence, against around a third in the US. UK recruiters are also less willing to take a candidate's word that they can do the job: 15%, against 23%. Both countries want the same fix. 92% in each would run an objective check before interview.
How many recruiters were surveyed for this report? 1,428 US talent professionals, fielded in August 2026, and 483 UK talent professionals, fielded September 10 to October 7, 2026. All were directly involved in sourcing or supporting talent acquisition in the previous 12 months.
Does the 2026 report cover the UK? Yes. The report covers 1,428 US and 483 UK talent professionals, surveyed with the same questionnaire, exclusions, and weighting, so the two countries can be compared directly. Each country is reported on its own base, and US and UK figures are never combined.