If you work with separating families, you already know the pattern. Communication between the parties is where the conflict lives, the standard advice is to move it onto a co-parenting app, and the party sending the hostile messages is exactly the party who won’t agree to use one. The tool sits empty while the hostility carries on.

FenceChat is built the other way around. It works on one side only. Your client gets a FenceChat email address, and a FenceChat number on the SMS plans. The other party sends to it like any normal email or phone. A language filter removes the hostility, the facts are forwarded to your client, and every original message is kept in their encrypted Archive, with the date and time it arrived. The other party installs nothing and has no account.

Because it runs on your client’s side, they can also shape it to their situation. Beyond a sensitivity setting, they can guide the filter from their own Archive, keeping anything the filter over-removed and filtering out anything hurtful that slipped through, so it fits the specific person they are dealing with. Those choices are private to their account, and whatever the settings, a threat is always flagged and shown as one plain line rather than softened.

Why that matters to you: Your client can act on their own. There is nothing to negotiate with the other side and no consent to obtain. By default the other party gets a one-time notice the first time they message, which your client can switch off if telling them could put them at risk. Replies your client sends through FenceChat still end with a short “Sent via FenceChat using AI” line. They can stay in contact, meet their obligations, and keep a complete record, all without the other party’s cooperation.

Articles for professionals

Plain-language articles written for the people who advise and support separating families, to help you understand FenceChat and explain it to clients. Read them, and request the Partner Pack when you are ready to introduce FenceChat in your practice.

Accountability and complaints handling

FenceChat is operated by ROHAJO Pty Ltd t/a FenceChat, an Australian company (ABN 94 699 691 640). So you can refer clients knowing how we’re held to account.

Questions practitioners ask

Does the other party have to agree to use FenceChat?

No. FenceChat works on your client’s side only. The other party keeps using ordinary email and text messages, and installs and signs up to nothing. See the one-sided difference.

If hostile words are removed, has my client lost evidence?

No. The original message is kept beside the filtered version, sealed into the Archive in order, and a check of an export shows any removal. See filtered, not erased.

Will it look as if my client is hiding something?

The other party gets a one-time notice that their messages are processed by AI and stored by FenceChat, and everything they send is kept as they wrote it. See what the other party is told.

What if the AI changes what was said?

The filter takes out hostile wording and keeps the facts, and the original is always kept. Pattern labels describe the wording of a message, not a person. See filtered, not erased.

Is it suitable where family violence is a concern?

FenceChat only filters messages sent to its own user and gives no one access to another person’s account or inbox. It is not a safety plan and won’t suit every situation. See where family violence is a concern.

Does it fit existing orders or a parenting plan?

To the other party, FenceChat is an ordinary email address or number. See communication terms in parenting plans and orders for the common terms and where it fits.

Will my client miss something urgent about the children?

Facts, times and arrangements are kept. If the filter can’t run for any reason, the message is delivered unfiltered with a note, rather than held back. See filtered, not erased.

How do I get the records, or subpoena FenceChat?

Your client can export their Archive from their Dashboard, and our information requests page sets out how we respond to subpoenas and certified copy requests. See getting message records into a matter.

Who can see the messages, and where are they kept?

Messages are encrypted per account and staff access is logged. Our main database is in Sydney, and overseas providers are listed in our privacy policy. See accountability and complaints handling.

Who pays?

Only the person using FenceChat. The other party pays nothing and signs up to nothing. See OurFamilyWizard or FenceChat.