If you have applied for several roles recently and heard nothing back from most of them, it is worth asking a different question than “what did I do wrong.” Increasingly, the honest answer might be: nothing. A piece of software decided you were not a match before a person ever read your CV, and that software may not be working as well as either you or the employer assumes.
Two stories from October 2026 make this worth stopping on. The UK’s Information Commissioner’s Office (ICO) has named recruitment an enforcement priority in its draft guidance on automated decision-making. A global survey from background-check company Checkr found that most hiring managers now believe AI-assisted candidate fraud is outpacing their ability to detect it. Read together, they say something useful: AI in hiring is not yet trustworthy on either side of the desk, and that changes what a smart job search looks like right now.
What the ICO’s guidance actually says
The ICO consulted on its draft guidance on automated decision-making and profiling between 31 March and 29 May 2026, with final guidance expected in summer 2026. The draft reflects the Data (Use and Access) Act 2025, which replaced the old near-total ban on significant automated decisions under UK GDPR with a safeguard-led approach instead. The strictest limits now mainly apply to decisions based on special-category data, such as health or ethnicity.
Alongside the draft, the ICO published a report on automated decision-making in recruitment specifically, built from engagement with more than 30 employers. Its headline finding is that many employers using screening or ranking tools do not realise those tools count as automated decision-making at all.
That matters, because the ICO’s guidance sets out real obligations once a tool is recognised as making or shaping a decision about a candidate: a lawful basis for using it, in most cases a data protection impact assessment, and crucially, meaningful human review. Rubber-stamping a system’s shortlist is not meaningful review. A reviewer needs the authority, the information and the time to actually overturn what the software suggests.
This is guidance, not new law, and it does not ban AI in hiring. What it does is signal where the regulator’s attention, and its audits, will go next. An employer that buys an off-the-shelf screening tool cannot hand its obligations to the vendor. It remains the one accountable for how that tool is used, and this applies to any organisation recruiting for UK roles or processing UK applicants’ data, wherever it is based.
Why this is landing now
The ICO’s focus on recruitment would be worth noting on its own. It becomes more pointed next to Checkr’s Hiring Hoax 2026 report, based on a survey of 3,000 hiring managers published in October 2026.
68% of hiring managers believe AI now helps candidates fake identities better than employers can detect, up from 62% last year.
60% say AI has made virtual interviews harder to trust, up from 51%.
40% have personally interviewed a candidate later found to be using a false identity.
65% have found candidates who misrepresented their skills or experience.
Only 23% feel confident their current process would catch a fraudulent candidate.
This is a US-focused survey, but the pattern will feel familiar on this side of the Atlantic too. The honest reading of these two stories together is not that AI should be avoided, by employers or candidates. It is that neither side of the hiring process currently trusts the tools doing the sorting, and that uncertainty is now visible enough that a regulator is stepping in. For a genuine candidate, that is oddly reassuring. The system you are up against is not as sharp or as fair as it might feel from the outside, and some of what looks like rejection is simply noise.
What this means for your application
None of this means applications are pointless, or that it is worth trying to “beat” a screening tool with keyword games. If anything, the more employers push back against unreliable screening and rising fraud, the more value sits with doing the opposite of what a bot would do. Four shifts are worth making.
Write for a human reviewer, not a keyword scanner
Tailor your CV to the role, but write it as if a careful, slightly tired person on the other side will read every line, because the ICO’s guidance is pushing employers to make sure that is exactly what happens. Specific, verifiable detail, a number, a named outcome, a tool you actually used, reads as real in a way that generic competency language does not.
Expect, and prepare for, a live element
If 60% of hiring managers say AI has made interviews harder to trust, expect more employers to add a live problem-solving task, a portfolio review or a short work sample, precisely to get past what a screen cannot verify. Go in ready to talk through your own work in real time, not just recite it.
Keep a clean record of what you submitted
With fraud concerns rising, you may be asked to confirm or demonstrate something you claimed. A simple record of which version of your CV, and which claims, went to which employer means you can speak to your own application confidently and consistently, without having to guess what you said six weeks ago.
Treat a silent rejection as data, not a verdict
An application that disappears without a reply may say more about an overloaded or poorly calibrated screening process than about you. It is worth noting where this keeps happening and adjusting your approach for that type of employer, rather than carrying it as a judgement on your ability.
What good use of AI in hiring can look like
None of this is an argument against AI in recruitment altogether. Used well, it can cut the time a genuine candidate waits for a first response, and free up a recruiter’s time for the parts of the process that need a person: the conversation, the judgement call, the context no form can capture. The problem the ICO and Checkr are both describing is not AI itself, it is AI used without real oversight, layered onto a system already struggling with volume. The practical test for job seekers is simple: does an employer’s process still involve a person who can explain, and change, a decision about you? If it does, you are dealing with the better version of this technology.
The bigger picture
Regulation is starting to catch up with the reality of AI in hiring, and that is a good thing, even though it will take time to bite. In the meantime, the candidates who do best are the ones who assume the process is imperfect on both sides and lean into what a screening tool cannot fake: specific evidence, real stories and the ability to explain your own work under a bit of scrutiny. That is also exactly what good interview preparation should be building towards, screening tool or not.
If you would like a second pair of eyes on whether your CV is set up to survive this kind of scrutiny, a CV Critique gives you a clear, practical read on that, along with a 30-minute follow-up call to talk through what to change.
You will also find a free Job Search Tracker and other resources on the Career Resources hub, worth using to keep your own record as you apply.
