LinkedIn launches AI-powered people search for natural language queries

Finding the right person on LinkedIn has always felt like a chore of mastering Boolean search and wrestling with clunky dropdown filters. If you didn't know the exact job title someone gave themselves, good luck ever finding them.

Now, LinkedIn is finally overhauling search with AI. Instead of typing an exact name, role, or company, users can simply describe who they need in plain English.

The search bar itself is even swapping out the generic "Search" prompt for "I’m looking for…" to encourage natural questions. You can ask for "healthcare investors with FDA experience," "Northwestern alumni working in entertainment marketing," or simply "Who in my network can help me understand the US work visa system?"

The feature is rolling out to LinkedIn Premium subscribers in the US first, with plans to expand globally and eventually reach all free users in the coming months.

Beyond keyword guessing

Over the last two years, LinkedIn has shoved AI into ad copy, post drafting, and learning modules. In May, it rolled out conversational job search. Bringing that same technology to people search targets the site's most critical utility.

LinkedIn's senior director of product management, Rohan Rajiv, noted that traditional lexical search forced users to know exact phrasing, leaving countless valuable connections buried. The new AI engine parses natural queries and ranks results based on personal graph proximity alongside thematic relevance.

Still, early testing shows it is far from bulletproof. Queries like "people who co-founded a YC startup" turn up noticeably different results than typing out "Y Combinator."

Even funnier, searching for "voice AI startup co-founders" surfaced people who simply hold a LinkedIn "Top Voice" badge. LinkedIn admitted it is still fine-tuning how the model interprets nuanced phrasing.

Why it matters

We think this is LinkedIn's most genuinely useful AI update to date. Generating automated comments or profile summaries always felt gimmicky, but eliminating the friction of networking and talent discovery solves a real headache.

For recruiters, founders, and professionals, this makes proactive outreach substantially faster, assuming you have US Premium access. It is worth trying today, though users should be mindful of early hallucinations and prompt quirks, like the amusing Top Voice mix-up.

For the broader AI ecosystem, this rollout highlights the enduring strength of proprietary data moats. Autonomous browsing agents and AI search startups frequently demo scraping LinkedIn profiles to build automated headhunting pipelines.

Yet Rajiv hinted that LinkedIn will likely harden its data policies over time, mirroring Reddit's crackdown on unpermitted AI scraping. When the platform owning the underlying identity graph builds competent conversational search natively, third-party scrapers lose their edge almost overnight.