AI counsel before you post
The mistakes that trip people up most often on social media, with a plain-English explanation of why they matter and the safer approach to use instead. Read this before you draft, and screen your post before you publish.
Ten recurring traps, grouped by the area of law they touch. Each one shows the mistake, why it's risky, and the safer way to say the same thing.
The mistake
Writing 'in my opinion, they're a fraud' doesn't make a defamatory statement safe. Courts look at whether the core claim is provably true, not whether you prefaced it with 'I think'.
Safer approach
Frame opinions around facts you can verify, or state the underlying facts plainly and let readers draw their own conclusion.
The mistake
You don't have to use a name. 'The manager at that Bondi cafe who always skims tips' can identify a person just as clearly, and the imputation of theft is defamatory.
Safer approach
If the subject could be identified by context, assume the post identifies them. Remove identifying details or make sure every claim is true and provable.
The mistake
A photo on Google Images or a competitor's feed is not free to use. Embedding or reposting someone else's image, video, or music without permission is infringement even with attribution.
Safer approach
Use licensed stock, Creative Commons material with the right licence, or your own original media. When in doubt, ask the rights holder.
The mistake
Screenshotting a news article, comic, or graphic to 'react' to it can still infringe, especially if your post substitutes for the original rather than commenting on it.
Safer approach
Quote or crop the minimal portion needed for genuine commentary, and link to the source rather than reproducing the whole thing.
The mistake
Using a brand's name, logo, or tagline in a way that suggests sponsorship or approval, even unintentionally, can be trademark misuse and misleading advertising.
Safer approach
Make clear when you have no relationship with a brand. Use disclaimers like 'not affiliated with' where there's any chance of confusion.
The mistake
In several jurisdictions, repeating or amplifying a defamatory statement by sharing, reposting, or quote-tweeting it counts as a fresh publication, you adopt the risk.
Safer approach
If the original is defamatory, add your own framing and verification rather than simply amplifying. Sharing is publishing.
The mistake
Private people have stronger privacy protections than public figures. Naming a colleague, customer, or bystander in a post, even a positive one, can breach privacy or data-protection law.
Safer approach
Get consent before naming or showing private individuals. If consent isn't possible, anonymise or omit the identifying details.
The mistake
Venting about a workplace, a client, or a project can breach confidentiality clauses, NDAs, or non-disparagement agreements, even if every word is true.
Safer approach
Check your employment contract and any NDAs first. Describe situations in general terms, and never disclose confidential information or client identities.
The mistake
Failing to clearly disclose paid promotions, gifted products, or affiliate links misleads consumers and breaches advertising standards and consumer-protection law.
Safer approach
Use clear labels like #ad, 'Paid partnership', or 'Affiliate link' at the start of the post, and disclose before any user has to expand the caption.
The mistake
Discussing an active court case, naming parties under suppression, or speculating about a verdict can amount to contempt of court or sub judice breach.
Safer approach
Before posting about a legal matter, check whether it's still before the courts and whether any names are suppressed. Report facts, speculate elsewhere.
Plain-English explainers that go deeper on each area of risk, plus our white paper on pre-publication screening.
What counts as a defamatory post, who can sue, and the defences that protect honest opinion.
4 min readWhere whistle-blowing ends and breach of contract or confidentiality begins.
4 min readReposting, screenshots, music, and the narrow limits of fair use.
4 min readPrivacy, data protection, and when naming a private person crosses the line.
4 min readHonest reviews, the protection of truth, and how to avoid a SLAPP threat.
4 min readOur white paper on why AI-assisted pre-flight checks reduce legal risk for creators and brands.
8 min readPaste your post and get a risk verdict, flagged issues, and a safer rewrite — free, anonymous, in seconds.
Run legal pre-flight