Quick answer
If you’ve posted a design or development role in the last year, you already know the drill: you get buried. A couple hundred applications land in a day, most of them clean, professional, and completely interchangeable. That’s not because the field suddenly got more qualified people in it. Applying just takes ten seconds now instead of twenty minutes. The hiring managers doing well right now aren’t trying to read more resumes faster. They’re testing for real work instead of polished writing, and sourcing from pools where the noise hasn’t gotten in yet.
What’s actually changed
Applying to a job used to take real effort. Now a candidate can run a posting through ChatGPT and have five tailored applications out the door before their coffee’s cold.
LinkedIn reported close to a 45 percent jump in applications submitted on its platform this year, hitting roughly 11,000 submissions a minute. The New York Times tied that surge directly to candidates using generative AI to draft and submit applications faster. A Financial Times investigation put AI use among applicants at around half, with recruiters describing applicant pools that more than doubled in size for a single role. Robert Walters’ latest survey found 70 percent of employers have seen a jump in applications because of AI, and about two-thirds of professionals now use AI tools specifically to apply to more roles at once.
SHRM’s 2026 hiring predictions put it plainly: HR teams are struggling to find real talent inside that flood, and the AI tools brought in to screen it sometimes make the problem worse.
For a hiring manager, that adds up to one thing. A single posting can pull in hundreds of well-written, similar-sounding applications, and almost none of them tell you whether the person behind it can actually do the job.
What doesn’t work anymore
- Posting broadly and hoping the right resume surfaces. It gets buried under everything else before you ever see it.
- Screening for resume polish. Everyone’s resume is polished now, so it doesn’t tell you anything.
- Generic screening questions like “why do you want to work here.” Ten seconds in ChatGPT and it’s answered, without teaching you a thing about the candidate.
- Assuming more applications means more choice. Mostly it just means more hours spent screening for the same handful of good candidates you’d have found anyway.
What works instead
Be specific enough to filter people out. Name the actual problem the role solves and the exact tools or stack involved. A vague posting invites everyone. A specific one invites people who’ve actually done comparable work.
Ask for something a candidate can’t outsource to AI. Skip “walk us through your experience” and ask for a short response to your actual product, or a small paid work sample instead. A generic cover letter takes ten seconds to generate. Five years of judgment on a real problem doesn’t.
Pay attention to process, not just the polish. Ask a designer how a decision got made, not just what the final screen looks like. Ask a developer to pair with you for twenty minutes instead of grading a take-home. You’ll learn more from watching someone think than from anything they submit.
Source from pools that are already self-selected. Mass job boards optimize for reach. Niche boards built around people who already work in design or development optimize for fit. A pool that’s already filtered by relevance beats one filtered by nothing.
Once you find someone real, move fast. Good candidates aren’t sitting around waiting. Running them through another two rounds “to be thorough” mostly just gives them time to accept something else.
Common mistakes
- Treating a flood of applications as a good problem. It costs screening time and buries the people worth talking to.
- Assuming an AI screening tool solves an AI-flooded funnel. Often it’s just two AI systems working against each other, one writing applications and one filtering them, with less human judgment in the loop than either side assumes.
- Weighing the tools listed on a resume over what someone actually built with them.
- Skipping a real work sample because it adds friction. That friction is doing useful work. It’s one of the few filters AI can’t fake its way through.
FAQ
Has AI actually increased the number of job applications? Yes. LinkedIn reported close to a 45 percent increase in submissions on its platform, and separate surveys from Robert Walters and Beamery put AI-assisted applications somewhere between 46 and 65 percent of job seekers, depending on the market surveyed.
Can I reliably tell if a resume was AI-written? Not from formatting alone. Recruiters interviewed by the Financial Times pointed to generic phrasing repeated across applicants as the clearest tell, but the more reliable fix is testing for real work rather than trying to detect AI in writing.
Should I use AI screening tools to manage the volume? They can help you triage, but the root problem stays put: generic postings attract generic applications. Narrowing the funnel earlier does more than filtering harder after the fact.
Where should I post open design and development roles? High-volume mass boards still make sense for less specialized hiring. For design and development specifically, boards built around an audience already working in the field tend to produce a shorter, more relevant list of applicants instead of a bigger one.
The takeaway
Applicant volume used to be the bottleneck in hiring. These days there’s no shortage of applications, just a shortage of ways to tell which ones are real. It’s part of why a niche, curated board is worth trying alongside the mass ones: Authentic Jobs skews toward an audience of designers and developers who showed up to do the work, not because a tool applied for them.
