“Our hiring got faster,” said the head of talent at a Bengaluru based unicorn. “But so did our fatigue.”
Across India’s startup hubs, that sentence has started to sound routine.
Eighteen months ago, AI hiring platforms were sold as the solution to everything broken in recruitment. The idea was simple: let algorithms do the grunt work, remove bias, and close roles faster. For a while, it looked like it worked. Dashboards looked clean, recruiters looked relieved, and companies congratulated themselves for being data-driven.
Then the numbers stopped telling the full story.
The Promise vs The Reality
At first, automation made life easier. But as one Mumbai fintech discovered, faster doesn’t always mean smarter. The company cut its average time-to-hire by nearly half after adopting an AI screening tool. The CFO called it a revolution. Six months later, HR realised that most of the new hires came from the same handful of colleges and backgrounds.
“We thought we were innovating,” the CHRO said. “Turns out, we were scaling sameness.”
Speed had improved. Diversity had disappeared. No one had budgeted for that outcome.
What AI Actually Broke
Cultural Drift
Most AI systems learn from past success stories. That sounds efficient until you realise they reward patterns, not possibilities. A Pune pharma firm found its AI recruiter consistently shortlisted candidates from the same universities as its senior leadership. Innovation was quietly filtered out before interviews even began.
Candidate Fatigue
Automation made the hiring funnel wider but colder. “It felt like applying to a vending machine,” a designer wrote in a feedback form. A Bengaluru startup tracked its offer acceptance rate and saw a drop of almost one-third among candidates who never interacted with a human during screening. The message was clear: people still want to be hired by people.
Data Debt
Every AI tool promises smarter hiring but leaves behind a trail of digital clutter — résumés half-parsed, duplicate profiles, untagged feedback forms, old video files that no one ever deletes. Over time, that mess becomes invisible friction. Recruiters waste hours cross-checking information across systems that don’t talk to each other.
Invisible Labour
Behind every “autonomous” system sits a team of recruiters quietly fixing its errors. A TA lead at a Series B SaaS firm laughed when she said her real title should be “AI Babysitter.” Her team spends almost a full day each week reviewing what the tool gets wrong and try to make he model learn about their hiring algorithms.
The New Problem: Decision Overload
Bias didn’t vanish. It just started wearing numbers.
Hiring managers no longer debate intuition; they argue over algorithmic confidence scores. A Hyderabad based recruiting manager described how he now receives a ranked list of thirty candidates, each tagged with skill-fit percentages and culture metrics. “I spend more time checking if the AI is right than I ever did choosing on my own,” he said.
The promise was less work. The result was information fatigue.
How Companies Are Adapting
Some HR leaders are pulling back from automation overload. Instead of chasing every new platform, they are focusing on making a few tools actually useful.
A Gurgaon based tech company reduced its AI stack from five vendors to two. Hiring slowed slightly, but satisfaction scores rose across recruiters and candidates. “We were drowning in automation,” their VP of People said. “Less really was more.”
Transparency is becoming another quiet trend. A handful of firms now disclose when and how AI is used during selection. Some even allow candidates to appeal automated rejections for a human review. What sounded risky on paper turned out to build trust.
The New Budget Line Items
The smartest CHROs have started budgeting for what they call “the human layer” around AI.
They fund bias audits run by outside experts, training programs to help recruiters understand when to override machine recommendations, and internal governance boards that treat model updates like policy changes.
We may treat AI deployment the way we treat a product launch. There’s a release plan, a rollback plan, and a feedback loop.
None of this shows up under technology costs. It’s all adaptation spending.
What’s Actually Working
AI will keep making hiring faster. That part isn’t changing. What matters is how companies use that speed. The ones seeing real progress are not chasing automation for its own sake. They are choosing where to keep humans in the loop and where to let data take over.
They have realised something that technology vendors rarely mention. The hardest part of hiring was never sorting résumés. It was exercising judgment and building relationships with candidates.
And that, no matter how advanced the algorithm, still belongs to people.
