LinkedIn scraping is the automated extraction of profile data (name, title, company, experience) with software instead of copying it by hand. It carries two separate risks: a contractual one, because LinkedIn's User Agreement prohibits bots and extensions that copy profiles, and a data protection one, because data visible on a profile is still personal data under the GDPR. Complying with one does not settle the other. The legitimate alternatives are manual search, LinkedIn Recruiter, the official API, the agency's own ATS database, referrals and sourcing tools that document the purpose and origin of every data point.
What is LinkedIn scraping
LinkedIn scraping is the automated collection of information from LinkedIn by a program that reads profiles and saves their fields into a database or a spreadsheet. Instead of reviewing profiles one by one, the recruiter runs an extractor that collects the name, headline, company, experience and location of hundreds of them.
LinkedIn scraping appeals to recruiters because the network concentrates professionals who are not looking for a job. According to the Statista survey on LinkedIn usage frequency in Spain from March 2024, 21% of respondents use LinkedIn daily and 24% once a week.
Scraping is different from two neighbouring practices:
- Manual search: the recruiter reviews profiles and notes only what they need.
- Sourcing tool: search, filtering and outreach in a workflow with defined access and processing.
- Uncontrolled scraping: piles up profiles without limiting fields, purpose or retention.
How LinkedIn scraping works step by step
A LinkedIn extractor turns pages designed for people into records. It does not need artificial intelligence: locating fields in the page code and saving them is enough.
- Define the search: filters or a list of profile URLs.
- Open each profile: with a browser and a logged-in session, which only sees what that account can see.
- Parse the HTML: if LinkedIn changes its layout, the extractor leaves fields empty or mixes up the previous and current employer without raising an error.
- Save the fields: to a CSV or a database, ideally with source and date.
- Repeat with pauses: high request rates and repetitive patterns trigger account restrictions.
Methods to extract LinkedIn data and their risk to the account
There are five ways to get data out of LinkedIn, and the risk to the account grows with unauthorised automation. Browser extensions and in-house scripts are the methods LinkedIn explicitly prohibits.
| Method | How it works | Indicative scale | Account risk |
|---|---|---|---|
| Manual copy | The recruiter copies what is visible | A few profiles a day | Low |
| Browser extension | Reads the page while you browse | Dozens to hundreds a day | High: prohibited by LinkedIn |
| In-house script (crawler) | Simulates browsing and extracts fields | Thousands of profiles | Very high |
| Third-party tool | Automates search and extraction | Thousands of profiles | Depends on the vendor |
| Official LinkedIn API | Access for approved partners | Approved use cases | Low, within its terms |
These scales are not safe or authorised limits. Account risk does not measure data protection risk either.
Is LinkedIn scraping legal under LinkedIn's terms of use?
No: LinkedIn's User Agreement prohibits scraping profiles. Its help page on prohibited software and extensions on LinkedIn points to section 8.2, which bans the use of "software, devices, scripts, robots or any other means or processes (such as crawlers, browser plugins and add-ons or any other technology) to scrape or copy the Services, including profiles and other data from the Services".
LinkedIn's section 8.2 also prohibits getting around usage limits on searches and profiles. It is a contractual ban, and its typical consequence is the restriction or closure of the account. A data protection legal basis does not lift it, and contractual permission does not replace a legal basis either.
Risks of LinkedIn scraping under the GDPR and Spain's AEPD
In the EU, data extracted from a LinkedIn profile is personal data under the GDPR (and, in Spain, the LOPDGDD) even when the profile is public. Even the scores and inferences an algorithm produces about a person can be personal data, as the Forlopd legal analysis of algorithmic systems in the workplace explains.
Spain's data protection authority, the AEPD, has already treated public LinkedIn profile data as personal data: in its decision on LinkedIn's 2018 security flaw it investigated how unapproved websites could harvest profile data through the "AutoFill" button, and closed the case after confirming the fix.
Before using LinkedIn data in a campaign, the agency must be able to justify:
- Legal basis for that specific processing.
- Purpose: which vacancy the data is collected for.
- Minimisation: why those fields and not others.
- Transparency and rights: what the person is told and how they can object.
Legitimate interest is not a universal permission: it requires weighing necessity, the person's reasonable expectations and their rights, and documenting that balance.
Which LinkedIn data is most sensitive for a recruitment agency
The closer a LinkedIn data point is to direct contact or to an algorithmic inference, the more the GDPR demands to process it.
| Data extracted | Nature | Main requirement | Risk |
|---|---|---|---|
| Name, title and company | Identifying | Legal basis and minimisation | Medium |
| Work history and education | Profiling value | Notice and retention period | Medium-high |
| Work email | Contact | Basis for contact and right to object | High |
| Mobile phone | Personal contact | Justified need | High |
| Inferred age or language | Algorithmic inference | Risk assessment and human oversight | Very high |
When the data does not come from the person, Article 14 of the GDPR as published in Spain's official gazette (BOE) requires informing them of its origin within a reasonable period of at most one month, or earlier if they are contacted. The guide to assessing a sourcing tool under the GDPR applies that analysis to the vendor.
AI filters and automated decisions in LinkedIn sourcing
An AI filter that scores LinkedIn candidates is profiling, and the GDPR restricts decisions based solely on that automated processing. Ranking candidates with AI is acceptable; rejecting them without real human review is not.
In 2026 the AEPD sent a warning to a company about its AI tool for screening applications, calling for data protection by design and an impact assessment where risk is high. In its note on safeguards when using AI to analyse CVs, the AEPD states that whoever decides "must be able to critically assess the information or score provided by the system and make their own decision" (our translation).
Ranking Java specialists by evidence of distributed systems work is prioritising. Rejecting anyone who seems to be over 50 based on their graduation date is a discriminatory criterion and has no place in the process. Log the filters used, the date and who reviewed. The difference between AI-assisted hiring and professional judgement lies in who keeps the final call.
Checklist for auditable LinkedIn sourcing
LinkedIn sourcing is auditable when the agency can say who approved the search, where each data point came from and what was done with it.
- Approved campaign: vacancy, owner and purpose before searching.
- Per-candidate record: source, date, legal basis and objection status.
- Minimum fields: only what affects the evaluation or the contact.
- Stop on restrictions: on blocks or CAPTCHAs, stop; do not work around them.
- Retention and deletion: a review date, exported spreadsheets included.
- Objections honoured: nobody who objected comes back in from another list.
Legal alternatives to LinkedIn scraping in 2026
There are six legitimate alternatives to LinkedIn scraping, and each one covers platform access, the legal basis for processing, or both. None of them legitimises every later use on its own.
| Alternative | What it provides | Access or processing basis | For whom |
|---|---|---|---|
| LinkedIn Recruiter Lite | Advanced search and InMails | Subscription under its terms | Individual recruiters |
| LinkedIn Recruiter Corporate | Search and team collaboration | Licence within the authorised scope | Large TA teams |
| Official LinkedIn API | Data in your own systems | Access approved by LinkedIn | Companies with in-house developers |
| Own ATS database | Talent already on file | Purpose compatible with collection | Agencies with history |
| Referrals and inbound | Candidates who apply or agree to an introduction | Contact requested by the person | Any search |
| Sourcing tool such as HeyTalent | LinkedIn search, AI filters, emails, phones and outreach | Public data under legitimate interest, according to the vendor | Agencies and headhunters with recurring searches |
The analysis of LinkedIn Recruiter alternatives in Spain ranks those options by need.
How HeyTalent fits as an alternative to LinkedIn scraping
HeyTalent is a sourcing platform for agencies and headhunters that combines LinkedIn profile extraction, custom AI variable filters, email and phone finding and LinkedIn outreach sequences, with no LinkedIn licence required. Connection campaigns with a note may require LinkedIn Premium Business.
On data protection, HeyTalent states that it processes public data under legitimate interest, transparently, and never resells data. The agency remains responsible for the purpose and legal basis of its own campaigns.
On data quality, HeyTalent states that its LinkedIn profiles are 100% up to date because they are extracted in real time at search time. That describes the method, not the accuracy of contact data.
On account risk, HeyTalent's experience with the accounts it manages puts the limit at roughly 100 invitations a week on a free account and up to about 200 on a paid one; LinkedIn publishes no official figure. HeyTalent stops sending when the account reaches its limit and pauses campaigns if an account gets close to a block.