Picture the scene: an employee lets off steam online after a bad day, forgets their bio still says where they work, and by Monday morning HR has a screenshot, an angry customer email, and a decision to make. Do you have the right to act? And if you dismiss them, will it hold up?
In Part One we looked at the post that ends in a lawsuit. This instalment is about the post that ends in a disciplinary hearing, a different question with a trickier answer. It isn't enough that a post is embarrassing, or even that it's true and unflattering. The law asks something narrower: does this post actually connect to the employment relationship, and if it does, is dismissal a fair response, or too heavy a hand?
Three cases, two of them decided this year, show exactly where our courts draw that line, and they don't all point the same way.
1. "It Was on My Personal Account" Isn't the Trump Card Employees Think It Is
The starting point is genuinely in the employee's favour: what someone does on their own time, on their own device, has nothing to do with their boss, as a rule. An employer doesn't get to police an employee's private life just because it employs them.
But there's a well-worn exception, and it's swallowing the rule fast: if the post connects back to the employer, the private-life defence falls away.
Edcon Limited v Cantamessa and Others [2019] ZALCJHB 273; [2020] 2 BLLR 186 (LC) is the case everyone in this space knows. A senior buyer, on leave, vented about the government on Facebook using language that was widely and reasonably read as racist. Her post had nothing to do with her job. The problem was her bio, which named her employer. A customer complained, Twitter piled on, a newspaper ran the story under the headline "Racist Monkey slur strikes again", and Edcon's name was in the middle of it. The CCMA initially sided with the employee, partly because Edcon couldn't point to a rand figure in lost business. The Labour Court wasn't having it: reputational risk is enough on its own. You don't need to prove the damage already happened, you need to show it plausibly could.
Takeaway for employers: if an employee's profile identifies them as yours, their "private" posts are a lot less private than they think. And you don't need a social media policy sitting in a drawer somewhere for this to apply, genuinely serious conduct speaks for itself.
2. Not Every Bad Post Deserves the Same Punishment
Here's where it gets more interesting, because our courts have made clear that "it's on Facebook" doesn't mean "automatic dismissal."
At the serious end: Dagane v SSSBC and Others [2018] ZALCJHB 114; [2018] 7 BLLR 669 (LC). A police officer posted comments online that crossed clearly into hate speech and incitement. Dismissal upheld, review dismissed with costs, no real debate to be had once language like that is on the table.
At the other end, more recently: Erarite (Pty) Ltd t/a Khayelitsha Superspar v CCMA and Others [2026] ZALCCT 53 (23 March 2026). A bakery manager, working somewhere ugly xenophobic tension had already flared, posted a Bible verse to his WhatsApp status that read as pointed given the timing. Badly judged, certainly, even the CCMA commissioner agreed with that. But not hate speech, not incitement, and nowhere near as serious as the campaign happening around him. The court agreed that a final written warning, not dismissal, was the fair outcome, and upheld three months' compensation for the employee. Gandidze J was also clear on something employers often get wrong on review: you're stuck with the case you actually argued the first time round. You don't get a second, tougher run at it later.
Takeaway for employers: ask yourself honestly where a post sits on the spectrum before you reach for dismissal. Hate speech and incitement, almost always fair game. A misjudged post in a tense moment, probably calls for a warning, not a P45.
3. If You're Going to Discipline Someone, Do the Homework First
Academy Brushware (Pty) Ltd v CCMA and Another [2026] ZALCCT 131 (12 August 2026) is a good cautionary tale, even though it's really a story about a shaky instruction rather than a shocking post. An employee with fifteen years' service posted photos of herself posing next to the company's own marketing cutout. The employer said she'd been told not to and did it anyway; she disputed ever getting that instruction clearly. The Labour Court's message: if you're relying on "we told them not to," you'd better be able to prove exactly what you told them, when, and how. Vague recollections don't cut it, and the review was dismissed.
Put simply, before you discipline anyone over a post:
Have a policy that actually says something about off-duty conduct and identifying with the company, not just "be nice online."
If you want something taken down, say so clearly, and put it in writing.
Charge the person for what they actually did, not a vaguer catch-all.
Be ready to point to real or likely harm, not just "this feels bad."
Treat similar cases the same way, nothing sinks a dismissal faster than someone finding a colleague who did the same thing and only got a warning.
On evidence: in Sedick and Another v Krisray (Pty) Ltd (2011) 8 BALR 879 (CCMA), the employer viewed posts on a wide-open profile, no privacy settings at all, and the CCMA said fair game, no invasion of privacy. The lesson for employers gathering evidence: public is public. But the moment you're accessing something genuinely private, a locked account, a private message, the ground shifts, and that evidence can become a liability rather than an asset. When in doubt, grab the screenshot and the timestamp immediately; posts vanish fast once someone realises they're in trouble.
4. This Cuts Both Ways
It's worth remembering that inaction can be just as risky as overreaction. The Code of Good Practice on the Prevention and Elimination of Harassment in the Workplace (18 March 2022, under the Employment Equity Act 55 of 1998) treats online harassment and cyber-bullying as squarely the employer's problem, even when it happens off the clock. If a staff member complains that a colleague is harassing them online and nothing is done about it, the business that ignored it can end up carrying its own liability, quite separately from whatever the harasser is facing personally.
The Bottom Line
Three questions, every time: Does the post genuinely connect to the job? How serious is it, really? And can you actually prove what you're relying on? Employers who can answer all three confidently are in good shape. Employers acting purely on outrage tend not to be.
Up next in this series: what happens when a post crosses from merely offensive into criminal territory. We look at the Cybercrimes Act 19 of 2020, where that line sits, and what actually happens when someone walks into a police station with a screenshot instead of walking into our offices.
This article is published for general information and does not constitute legal advice. For advice on a particular matter, please contact Gittins Attorneys Inc.
.jpg)
