Defamation and Course Evaluation: When Open-Text Comments Become a Libel Risk

Bottom line up front: A student's free-text comment that a lecturer "is disorganised" is an opinion, and opinions are strongly protected. A comment that the same lecturer "turned up to the 9 a.m. drunk" or "stole a student's work" is a statement of fact — and if it is false and your institution reproduces it in an evaluation report circulated to a head of department, a promotion panel or a programme review, your institution has published it. Publication is the act that creates defamation liability, and "we only passed on what a student wrote" is not, by itself, a defence. Most evaluation systems forward open-text comments verbatim and unread. That is a small but genuine legal exposure — and, more importantly, a fixable one.

This is not an argument for censoring criticism. Robust, even harsh, student feedback is legitimate and legally safe. The narrow problem is the false factual allegation, and the fix is a filtering step that most legacy tools simply do not have.

Opinion is protected; false facts are not

English law, reformed by the Defamation Act 2013, is a useful reference point because it draws the line cleanly — and many European systems draw a comparable, if differently worded, distinction. A defamatory statement is one that would cause an ordinary person to think worse of the claimant. Since the 2013 Act, a claim also requires proof of serious harm to reputation under Section 1 — a deliberately high bar that filters out trivial gripes.

Two defences matter here. Section 2 provides a complete defence of truth: if the sting of the allegation is substantially true, there is no liability. Section 3 provides the defence of honest opinion: a statement is protected if it was a genuinely held opinion, indicated at least in general terms the basis for it, and was one an honest person could hold on the known facts. "This module was badly organised and I learned little" is quintessential honest opinion. It is safe.

The exposure lives in the residue: a false statement of fact about identifiable conduct — criminality, dishonesty, intoxication, discrimination — that is neither true nor mere opinion. Those are the comments that, once circulated, can meet the serious-harm threshold, especially where they touch an academic's professional integrity.

The institution is a publisher

The word that does the legal work is publication. In defamation, every repetition is a fresh publication, and the person who repeats a libel is generally liable for it as if they had originated it. When a quality office takes a student's free-text comment and reproduces it in a report that reaches a third party — a line manager, an appointments committee, an external reviewer — the institution has published the statement to that audience.

Internal circulation may attract qualified privilege: communications made without malice between people with a shared duty or interest — for instance, a confidential report to the manager responsible for a member of staff — are protected. But qualified privilege is fragile. It can be lost if the material is circulated more widely than the duty requires, or if it is handled with reckless indifference to its truth. An unfiltered comment forwarded to a wide distribution list, or leaked beyond the intended readership, is exactly the scenario in which the privilege frays.

On the Continent the framing differs but the risk is not absent: in Germany, for instance, defamation is addressed through the Criminal Code (Beleidigung, üble Nachrede and Verleumdung under §§185–187 StGB) alongside civil personality-rights protection, and a false factual assertion damaging professional reputation can engage both. The vocabulary changes across borders; the underlying exposure — false factual allegations, republished by an institution — travels.

"Isn't this just a licence to suppress criticism?" — the objection

The strongest objection is that a defamation lens will be used to sand the edges off legitimate feedback — that nervous administrators will redact anything unflattering and call it "libel management," hollowing out the honesty the exercise depends on.

The objection is right about the risk and wrong about the remedy. The answer is not to redact criticism; it is to redact false factual allegations while preserving the critical opinion. "The lecturer was hostile and dismissive" is protected opinion and must survive. "The lecturer assaulted a student" is a factual allegation of a crime that, if unsubstantiated, should never be forwarded verbatim to a promotion panel — it should be routed to the process that actually investigates such claims (safeguarding, HR, or the police), not laundered through a teaching-quality report. Distinguishing the two is a triage problem, not a censorship problem. Done well, it protects the accused academic, the accusing student, and the integrity of the feedback channel.

Where Koji fits

Triaging thousands of open-text comments for the rare defamatory factual allegation is exactly the task humans do badly at scale and AI does reliably.

Automatic thematic analysis with a moderation layer. Koji's analysis of open-text feedback can flag comments that make specific factual allegations about identifiable conduct — as distinct from evaluative opinion — so a human reviews them before they enter a circulated report.
Route, don't republish. A serious allegation surfaced by the system can be directed to the correct process rather than forwarded in a teaching report — the same "right channel" logic that governs safeguarding disclosures and whistleblowing content that arrives through the wrong door.
Conversational probing reduces bald allegations. Because Koji's AI moderator can ask a follow-up — what specifically happened? — vague, potentially defamatory assertions are often either substantiated into something actionable or resolved into evaluative feedback, rather than landing as an unexamined one-line accusation.
Consistent, documented handling. Standardized moderation gives you an auditable account of how flagged content was handled — useful if a decision to redact or route is ever challenged.

Koji does not give legal advice, and none of this eliminates the underlying duty to handle serious allegations properly. What it does is put a filter where legacy tools have a firehose. The same moderation engine powers the general research platform at koji.so, where open-text moderation matters just as much for customer feedback.

The practical takeaway

Stop forwarding open-text comments verbatim and unread. Build a triage step — automated flag, human decision — between collection and circulation. Preserve opinion ruthlessly; route serious factual allegations to the process designed to investigate them; and keep a record of how you did both. That single change closes most of the exposure while making the feedback more useful, not less.

Frequently asked questions

Can a university really be sued for a comment a student wrote? Potentially, yes. In defamation, whoever republishes a statement can be liable for it. When an institution reproduces a student's comment in a report circulated to others, it publishes that statement. Passing it on "as received" is not automatically a defence — though internal, confidential, malice-free circulation may attract qualified privilege.

Are negative or harsh comments defamatory? Usually not. Genuinely held opinion — that a course was poorly organised, or a lecturer unengaging — is strongly protected (the honest-opinion defence in England; comparable protections elsewhere). The narrow risk is a false statement of fact about identifiable conduct: criminality, dishonesty, intoxication, discrimination. Those are what a filtering step should catch.

What is the difference between fact and opinion here? "Boring", "disorganised", "unfair marking" are evaluative opinions. "Came to class drunk", "faked the results", "is a racist" assert facts about what happened. Opinion is defended as opinion; a false factual allegation is not. The distinction is the whole game.

Should we just delete anything negative to be safe? No — that would destroy the value of the exercise and is not what the law requires. Preserve critical opinion. Only false factual allegations need special handling, and the right response is usually to route them to the correct process, not to erase them.

Does anonymity change the analysis? Not for the institution's exposure. Whether or not the author is known, the institution that republishes the statement is a publisher. Anonymity may make the original author hard to pursue, which can concentrate attention on the republisher.

Is this only a UK issue? No. The Defamation Act 2013 gives a clean vocabulary, but most European jurisdictions protect reputation — often through both civil personality rights and, as in Germany, the criminal code. False factual allegations republished by an institution carry risk across borders; only the labels differ.

Koji for Education puts an AI moderation layer between raw open-text feedback and the reports your committees read — flagging the rare factual allegation while preserving every honest opinion. See how it works.