On November 13, 2025, professional networking platform LinkedIn launched a new AI‑powered 'people search' feature, initially rolling it out in batches to paid members in the United States. LinkedIn said the feature lets users use natural language to quickly find professionals on the platform who meet specified criteria, rather than matching profiles one by one through traditional keywords and cumbersome filters.
In demos, LinkedIn showed users can enter complex queries such as 'people in San Francisco with a product manager background and at least two years of experience in AI.' The system returns people cards ranked by match score; the cards do more than list names and titles, highlighting skill snippets relevant to the query, recent employers, the number of shared connections, and available contact routes (such as InMail or introductions via shared contacts). The feature combines large language models with vector retrieval: LinkedIn converts user input into semantic vectors to match skills and experience fragments in profiles, then uses a generative model to organize result summaries and suggest initial outreach messages.
TechCrunch noted the platform also offers preset filters (company, location, function, years of experience, etc.) to further refine results beyond the natural language query. LinkedIn emphasized the feature is not intended to replace its existing recruiting products, but to embed AI search capabilities into members' everyday workflows, helping HR professionals, hiring managers and proactive networkers locate targets more quickly. The initial rollout is available to Premium subscribers in the U.S., with the company planning to expand access gradually after evaluating feedback and performance metrics.
On privacy and visibility, LinkedIn said the platform continues to honor existing discovery settings: users can control how searchable their profiles are and the visibility of contact information. LinkedIn also noted that AI ranking and recommendations take into account public information and privacy preferences and will not override users' visibility settings. TechCrunch also cautioned readers about limitations and potential biases of AI search: model training and retrieval logic may amplify visibility differences across certain skills, industries or regions, causing some qualified candidates to be undervalued or overlooked. LinkedIn said it will continue investing in model monitoring and iteration, improving diversity and fairness metrics, and providing more transparency for recruiters.
For early‑career professionals and recruiters, the feature's appeal lies in speed and expressive simplicity — there is no need to break complex requirements into multiple compound filters; a single sentence can yield a structured, actionable candidate list. LinkedIn's generated outreach suggestions can also lower the psychological barrier to initial contact.
The AI people search combines LinkedIn's existing profile data with generative and retrieval AI, testing first with U.S. paid users. It could shift the starting point of recruiting from keyword searches to semantic discovery. However, whether the feature can maintain accuracy, fairness and user privacy at scale remains to be seen.