To keep UK social media background checks compliant, you need three things:

• You must obtain the candidate's informed consent before any screening happens
• Checks must only cover publicly available content, and everything needs to comply with UK GDPR
• The data you collect should be relevant and limited to what's necessary for the specific hiring decision

Having a written social media policy, one that defines what types of content would affect a decision and applies it consistently, is strongly recommended before running checks at scale.

Zinc's social media checks are designed to meet these requirements. Candidates give explicit consent as part of the standard Zinc journey, and the platform only analyses public posts. It never accesses private accounts.

The FCA introduced annual open-source internet searches as a Good Practice recommendation for regulated firms in 2018. SMCR Non-Financial Misconduct guidance, finalised in December 2025 and in effect from 1 September 2026, makes this type of screening increasingly relevant for financial services firms. In schools, KCSIE guidance updated in September 2025 now says schools 'should consider doing' online searches, and that candidates should be advised that online searches may be conducted.

In the legal sector, SRA guidance on offensive communications makes clear that a solicitor's personal social media activity falls within scope of the conduct rules, and that the reputational and financial risk this creates for the firm sits with the partner or senior manager responsible for compliance at every SRA-regulated firm.

Law firms and solicitor's practices are required to have systems and controls in place to identify, monitor and manage all material risks to the business, and the SRA warns that they may be asked to demonstrate compliance if a complaint arises. Social media screening, both at the point of hire and on an ongoing basis, is the logical control under that framework.

With companies moving away from character-based references in favour of fact-based references, social media checks are the best way to check a candidate's character outside of a traditional character reference or professional reference with character questions.

When layered together, social media checks and reference checks can give you a fuller picture of a candidate’s behaviour in the context of their work history.

No. Zinc only analyses public posts, re-shares, and likes. It never accesses private accounts, direct messages, expired content like deleted posts or expired stories, or any content that requires a login to view. This isn't just a policy position. It's a legal requirement under UK GDPR, and it's the right approach for candidate trust. Candidates are informed that publicly visible content will be reviewed as part of their screening.

Zinc checks publicly visible activity on Facebook, Instagram, LinkedIn, TikTok, X, Reddit, and Pinterest. Before the report reaches you, a human analyst reviews all AI-flagged content to remove false positives.

One thing to know: if candidates provide a work email address rather than a personal one during the check, some social media accounts linked to their personal email may not be identified. Using a personal email improves account coverage.

Zinc's social media checks scan for content across 11 risk categories: Disparaging content, Drug images, Drug or alcohol mentions, Violence, Gore, Nudity images, Discriminatory language, Profanity, Self harm, Sexually suggestive content, and Threatening behaviour.

Your team's review against your own social media policy is what determines whether something is actually relevant to your hiring decision.

Zinc covers the last five years of public activity. AI categorises flagged content first, then a human analyst reviews it to remove false positives before the report reaches your dashboard. The final decision always stays with you. If you don't have a social media policy yet, Zinc can point you toward guidance on what a good one looks like. It's worth having in place before you run checks consistently.

Bias prevention is built into the process at two levels:

Natural language processing categorises flagged content by behaviour type, not by demographic signals. The system is designed to surface concerning conduct, not personal characteristics. Protected characteristics aren't flagged as risk factors within the 11 categories.

A human analyst then reviews all AI-flagged content before it reaches your dashboard, specifically to filter out false positives and content related to protected characteristics rather than genuinely concerning behaviour. When your team reviews results, decisions should be made objectively against documented criteria, not personal judgement.

Zinc's social media checks typically return results in 24 to 48 hours. That reflects the human analyst review stage. After AI has scanned and categorised public content across all relevant platforms, an analyst reviews the flagged material to remove false positives before the report is finalised. That step is what makes the results reliable enough to act on.

Candidates are asked for their full name, date of birth, country, and email address. Zinc uses this to identify their public profiles. After that, the check runs without any further input from your team or the candidate.