What can social media screening legitimately flag?

By HRjoyLast checked

Short answerGuide

A social media check can legitimately flag only what is relevant to the specific risk you documented for the role before anyone looked. Our view: in practice that means public evidence of conduct the role cannot accommodate, such as threats of violence, harassment or abuse, or disclosure of confidential information; public information that contradicts something the candidate told you; and, for roles working with children or adults at risk of harm, conduct that raises a safeguarding concern. The ICO's draft vetting guidance names these kinds of risk, including danger to others, unsuitable people working with children or adults at risk, and disclosure of trade secrets.

What a check must not flag is who the candidate is, what they lawfully believe, or a private life unrelated to the role. In Great Britain the Equality Act 2010 protects nine characteristics in recruitment, and the ICO says you should not make assumptions about suitability based on special category information such as race, political views or sexual orientation, even if the person made it public. Write the criteria down, give them to a reviewer who is not the decision maker, and filter everything else out.

Start from the risk you wrote down

Our main page lists the ICO's examples of risks that can justify vetting, from danger to others to disclosure of trade secrets. The Information Commission (ICO) says its draft recruitment guidance is under review following the Data (Use and Access) Act 2025.

Each flag criterion should trace back to a risk you documented for the role. If a finding does not connect to that risk, it is not a flag, however interesting it looks. This mirrors the ICO's care home example, where the person doing the checks screens the information for relevancy against the specified interview criteria.

Findings that can be job relevant

Our view: none should reach a decision maker until the candidate has had a chance to respond, which the ICO says you should offer.

Our view: findings that may be relevant, when, and the official source each draws on
FindingWhen it can matterBasis
Threats of violenceAny role where it shows a danger to colleagues, customers or the publicICO names danger to others as a vetting risk; for staff assessed under FIT, FCA FIT 1.3.21G gives threats of violence as an example
Harassment or discriminatory abuse of othersRoles with responsibility for colleagues, customers or people in their careFCA FIT 1.3.21G refers to conduct showing a material risk of harassment; ICO's charity example covers behaviour that would compromise the person's ability to do the job
Disclosure of confidential or commercially sensitive informationRoles with access to the same kind of informationICO names disclosure of trade secrets or other commercially sensitive information
Conduct raising a safeguarding concernRoles working with children or adults at risk of harmICO names employing unsuitable people in these roles; its care home example checks for conduct making someone unsuitable to work with people who need care and support
Contradiction of what the candidate told youAny role, if the point matters to the decisionICO sets out a process for discrepancies and says you cannot assume the candidate is being untruthful
Apparent involvement in criminal activityOnly where relevant to the role, and handled as criminal offence dataICO says criminal offence data can cover suspicion or allegations of criminal activity

What must not drive the decision

In Great Britain, section 39 of the Equality Act 2010 says an employer must not discriminate in the arrangements it makes for deciding to whom to offer employment, or by not offering employment. Section 4 lists the nine protected characteristics; our guide to protected characteristics and social media screening covers keeping them out of the process.

Political views need care. Acas says tribunals have found membership of a political party not to be protected, but that political beliefs could still be protected if they meet the definition of philosophical belief. Political opinions are also special category data, and the ICO lists political views among the things you should not make assumptions about. In Northern Ireland, nidirect says it is against the law to discriminate because of political opinion, though a political opinion which includes approval or acceptance of the use of violence for political purposes in Northern Ireland is excluded from protection.

Trade union membership is protected separately. In Great Britain, section 137 of the Trade Union and Labour Relations (Consolidation) Act 1992 makes it unlawful to refuse a person employment because they are, or are not, a member of a trade union. That Act does not extend to Northern Ireland.

  • anything revealing a protected characteristic, including what a photo suggests
  • lawful political, religious or philosophical views, or the absence of a religion or belief
  • trade union membership or activity
  • health, disability, sex life or sexual orientation
  • family, relationships, friends and who the candidate is connected to
  • lawful leisure activities and opinions unrelated to the role
  • old posts that do not suggest a current risk

Regulated roles in financial services

For staff assessed under the FCA's fit and proper test, FIT 1.3.21G says social media activity in private life is relevant where it indicates a material risk that the person will breach the requirements and standards of the regulatory system. Where it does not, the guidance says that, subject to FIT 1.3.17G, the activity is unlikely to be relevant and nothing in FIT requires a firm to concern itself with it. FIT 1.3.17G covers private conduct that shows a willingness to disregard ethical or legal obligations, abuse a position of trust or exploit the vulnerabilities of others, or that is serious enough to undermine public confidence in the regulatory system.

FIT 1.3.16G adds two limits: it should not be assumed that simply because a person engages in conduct in their private life there is a material risk they will repeat it in their regulated role, and a remote or speculative risk is not sufficient. Our view: both are worth applying to any criteria list, regulated or not.

Illustrative flag criteria you can adapt

This is an illustrative template, not a legal standard. Delete any line the role's documented risk does not support, and add the risk reference to each one you keep. Give the reviewer the final list in writing before the search starts.

  • Public statements threatening violence against a person or group (risk: danger to others)
  • Public posts harassing, abusing or demeaning individuals, including because of a protected characteristic (risk: harm to colleagues, customers or people in care)
  • Publication of an employer's or client's confidential information (risk: disclosure of commercially sensitive information)
  • Public content suggesting a risk to children or adults at risk of harm (risk: safeguarding, for roles working with those groups)
  • Public information contradicting the qualifications, employment history or other facts the candidate gave you (risk: reliance on false information)
  • For public-facing roles only: public conduct that would compromise the person's ability to do the job, described in specific terms rather than as a values match

Testing a flag before it reaches the panel

Our main page explains why online information cannot be relied on alone. In the ICO's political party example, the information might be out of date or inaccurate, the candidate might have been mistaken for another person with the same name, or they may have since changed their views.

Before passing a flag on, the reviewer should record what was found, where, its date, which criterion it meets and how they confirmed it relates to the candidate. The candidate then gets a chance to comment. Our false positives guide covers identity checks.

Sources

General information for employers in the UK, not legal advice. HRjoy advises on screening policy and fair process. We do not run checks ourselves; where you need checks run, we can introduce you to our screening partners. Check the official sources for your situation.

Questions

Something else? Ask us.

Can we reject a candidate whose public views we disagree with?

Not on that basis alone. The ICO says you should not make assumptions about suitability based on special category information such as political views, even where the person made it public. The narrow exception is information significantly relevant to the role, as in its example of a political party recruiting a secretary for its leader. For staff assessed under the FCA's fit and proper test, FIT 1.3.21G says that, subject to the rest of FIT 1.3, a person can lawfully express controversial views on social media without calling their fitness into question.

Does an old post count?

Only if it still suggests a current risk relevant to the role. The ICO warns that online information may be out of date, and that candidates may have changed their views. Record the date of anything flagged and ask the candidate about it before relying on it.

What if the reviewer sees something serious that is not on the criteria list?

Our view: if it suggests an immediate risk of harm to someone, the reviewer should raise it with whoever handles safeguarding or security concerns, not with the hiring panel. Otherwise it stays out of the decision, and you review the criteria before the next campaign.

Can a 'values' mismatch be a flag?

Only narrowly. In the ICO's charity example, a public-facing role leads the charity to consider checking that public behaviour is not inconsistent with its values, but the ICO says the charity must decide whether, on balance, the checks are necessary and proportionate. Our view: describe the conduct you are looking for, not a general fit.

Related checks

Sector guides

Guides

Related answers

All screening guidance

Which checks do your roles justify?

Tell us the roles you are hiring for. We’ll help you set a proportionate screening policy and candidate notices, introduce you to our screening partners when checks need running, and help with a fair process when a check comes back with something on it.

Ask HRjoy