Not satisfied with the court process

itsupportguy

New member
I am in a divorce battle with my spouse and 2 children are involved.

I have done one case conference and meditator session which failed and the meditator said I don't think we can come to any conclusion. I feel like I am not being heard because all my arguments and rationality is heavily evidence based.

In the first case conference the judge basically just give general ideas of how things go, support calculator information etc. Meditation is also same, what I am claiming is heavily based on rationality and evidence but I feel like these people are busy discussing general guidelines.

I do specific conversation but I feel like I am not being heard .

Feels like I need to change my strategy. The person I am dealing with is undiagnosed BPD and her unstable sense of self makes the entire forms look weird because it feels like they are filled by 2 different persons. Even ChatGPT can see all the contradictions but maybe we only get a chance to discuss this at trail ?​

Another thing:
I have often read about Section 30 assessment is useful but I dont think so because section 30 could be mainly analyzing a person in a given time but the inconsistencies spouse shows is based on weeks, months so she will be easily able to manipulate it. In such cases its better I have a medical professional at trail who can comment on the evidence I present e.g. unstable sense of self , how it impacts the child etc.
 
Your post is a little vague so it is hard to comment on why you are in the situation you are and how things could be different.

Yes the family court process in Canada is challenging as it is slow and can be unfair however, a lot of people have expectations for things should go and when it doesn't go the way they think it should, it is therefore bad.

You may need to reassess your approach and expectations. Yes evidence is good but it may not apply to your situation. For example, if your kids are older, a section 30 assessment may not be effective.

Not everyone has a mental disorder and Tayken has a great link related to people who assume their ex is BP or BPD. You thinking they are mentally unstable is not a diagnosis and you not getting what you want doesn't mean they have one.

It might be a good idea to take a step back and look at your case, your situation. You may also need to look at potential settlement options and what you would be willing to accept. If you don't have a lawyer, perhaps talking to one may help you focus a bit more on what is happening and what may happen.

Just some thoughts.
 
Yes, as I mentioned, there are children involved. I’m not overly concerned with a diagnosis like BPD per se—what matters to a trial judge is the current impact on the children. A diagnosis is just a label, if a parent can manage it, it doesn’t necessarily cause harm. The real issue is the problematic symptoms (evidence ) the spouse displays e.g. an unstable sense of self that research shows can lead to disorganized attachment in children (as supported by attachment theory). It seems that only at trial will I have the opportunity to present a fully evidence-based argument, as earlier processes tend to overlook these nuances.
 
Again, you should look up Tayken's posts about mental disorders. It doesn't matter what evidence you have about what could happen with someone who has a mental disorder. It would appear you are trying to argue a what if instead of arguing the right now. If there is no diagnosis or even an issue with the kids then you are blowing your chances and going all the way to trial is ridiculous especially since you could end up paying costs.
 
Again, you should look up Tayken's posts about mental disorders. It doesn't matter what evidence you have about what could happen with someone who has a mental disorder. It would appear you are trying to argue a what if instead of arguing the right now. If there is no diagnosis or even an issue with the kids then you are blowing your chances and going all the way to trial is ridiculous especially since you could end up paying costs.

All correct and excellent advice from Rockscan.

Some other great things to consider:

1. Forget Harmony, Settle for Peace
https://www.yoursocialworker.com/s-articles/Forget-Harmony.htm

2. Who important is consistency really?
https://www.yoursocialworker.com/s-articles/consistency.htm

3. Credible data in custody and access disputes
https://www.yoursocialworker.com/s-articles/credible_data.htm

4. Let go to win
https://www.yoursocialworker.com/s-articles/monkey.htm

5. A funny thing happened on the way to trial
https://www.yoursocialworker.com/s-articles/choice_of_assessor.htm

6. Who started it doesn't necessarily matter!
https://www.yoursocialworker.com/s-articles/circular_causality.htm

7. Mediating child beahviour problems between seperated parents
https://www.yoursocialworker.com/s-articles/cbsp.htm

I see most of your problem, OP, as requiring therapeutic intervention not legal or court intervention.
 
I am in a divorce battle with my spouse and 2 children are involved.

This is a huge problem. If you take the position that you are in a "battle" you are at war. No one wins in "divorce". Especially children will be harmed in "battle". I highly recommend you seek theraputic and legal counsel asap.

I have done one case conference and meditator session which failed and the meditator said I don't think we can come to any conclusion. I feel like I am not being heard because all my arguments and rationality is heavily evidence based.

Evidence based... As you are ITSupportGuy by name its clear you are an analytical thinker. Have you considered that your emotional state may be dirtying the data that you are calling "evidence based" and "rational". Evidence management is not clear cut and not a if-then-else function set. Its not a CASE STATEMENT. Evidence is weighted on the balance of probability and depends on critical factors of relevancy and cogency. You are in an emotional state where your evaluation of evidence, on the balance of probabilities, relevancy and cogent evidence is SIGNIFICANYLY deminished. You are a parent first, litigant second and if you are unrepresented by council you are also your own lawyer.

It takes SIGNIFICANT fortitude to be a parent, lawyer and litigant all at the same time. This is why the court system is backed up significantly. Because very few people can play this role and bring a matter to resolution. It is hard to even do when parents can see mostly eye-to-eye. When you start engaging into conflict everyone is looking at the sky and no one is looking at each other. This elongates the resolution to the problem significantly.

You are also looking for the courts to "fix" your problem but, it is your problem. You need to consider a wider viewpoint on how to resolve the issues and move everything forward for the sake of the children.

IIn the first case conference the judge basically just give general ideas of how things go, support calculator information etc. Meditation is also same, what I am claiming is heavily based on rationality and evidence but I feel like these people are busy discussing general guidelines.

Did you by chance look at that judge's docket for the day. I highly recommend you understand the time that is allocated to a case conference. Often parents waste this time trying to "argue" points that can only be done on motion. They often come out saying they were not being "heard" but, the reality is... Nothing but reasonable settlement happens at a Case Conference or Mediation. Nothing can be ordered. Only a MOTION, TRIAL or ARBITRATION results in an ORDER directing BOTH parents.

I do specific conversation but I feel like I am not being heard

Have you considered the following when you say "I feel like I am not being heard". A very wise judge in a court room once and has always stuck with me.

"Oh I hear you Mx. Person. What you are failing to understand is that I do not agree with you."

Lots of people rely upon this fallacy that stating "I am not being heard" as a project against the other party to the matter that they are not "listening". It poses the question to the person making the claim... Are you using this to try and settle, in your mind, an excuse as to why you didn't get your way? High conflict people often use the "I am not being heard" statement because, in "internet psychology" (not real!) the simple declaration of "not being heard" often is used as a weapon to call the other party a "narcissist". I note that this well documented behaviour pattern is occurring in your messages. Take a deep look at what you wrote and challenge yourself back on what is happening in the now and not projecting what "could" happen in the future.

Feels like I need to change my strategy. The person I am dealing with is undiagnosed BPD and her unstable sense of self makes the entire forms look weird because it feels like they are filled by 2 different persons. Even ChatGPT can see all the contradictions but maybe we only get a chance to discuss this at trail ?​

If you are going to use ChatGTP some recommendations:

1. Load your whole case file. You will need to include all communications, emails, TXT, SMS, affidavits, briefs, etc... loaded
2. THe scope for which you are evaluating is very small and you will need to point the learning model at CANLII with specific direction to evaluate all relevant caselaw.
3. ChatGTP won't solve problems but, it can assist in your analysis. It is not magic but, if you know how to instruct it you can get some good insights and may be useful in replying to anything in a non-confrontational manner.
4. Use it as a tool to help you better communicate and narrow the complexity of your "evidence". DO NOT USE ANY RESULTS AS EVIDENCE. The results and insights should help you train yourself and respond better to all of this. It won't be your free lawyer or show you a "slam dunk" response.

ChatGTP is a tool... Like this forum but, it will never provide you the answer.

Another thing:
I have often read about Section 30 assessment is useful but I dont think so because section 30 could be mainly analyzing a person in a given time but the inconsistencies spouse shows is based on weeks, months so she will be easily able to manipulate it. In such cases its better I have a medical professional at trail who can comment on the evidence I present e.g. unstable sense of self , how it impacts the child etc.

Section 30 assessments are complex and rarely yield positive results for either parent or the children. They are between 30-100k depending on where you reside. They have fallen out of favour with the superior court. All they usually reveal is the unreasonable party and voluminous affidavits already provide enough on parental behaviour to sort that out. As well, very few clinicians are willing to do them anymore and some have been disciplined for their conduct in them such as Dr. Peter Sutton in Kagan v Sutton, 2022 CanLII 94939 (ON HPARB) found here: https://canlii.ca/t/jsdx8

Para 93. states as a result of Dr. Peter Sutton (the Respondent):

93. Pursuant to section 35(1) of the Code, the Board confirms the Committee’s decision to issue advice to the Respondent as follows:

1) when conducting a child custody and access assessment, to make a thorough effort to follow up on all credible indications that a parent/spouse has shown a tendency to dishonesty or violence;

2) to consider the potential role of obtaining additional psychiatric assessment as part of the custody and access assessment; and

3) to consider interviewing a wide range of interviewees regarding both parents and the child when conducting child custody and access expert assessments.


It is a very complex case but, from my deep review of the whole unfortunate situation for Kegan was that there may have been a belief or hope that having a clinician do a Section 30 would have easily demonstrated the other parent's pattern of bad behaviour. I don't doubt that the mother, medical doctor as well, would have ever thought that matters would turn out as they unfortunately did, as another clinician was supposed to be evaluating the whole of the situation.

But, as para-93 demonstrates... This "section 30" evaluator needed an order to do their job persuant to section 35(1) of the Code of their governing practice.

So, I would recommend avoiding a Section 30 and the OCL as much as possible. They are not a tool that will win you anything. They are not a silver bullet and even "qualified" custody and access evaluators who are medical doctors need "advice" on how to do them.
 
Be careful with ChatGPT. The hallucination rate is relatively high (apparently 58% for GPT4, and higher for earlier models). Citation: https://academic.oup.com/jla/article/16/1/64/7699227
Hi @OrleansLawyer,

Appreciate the cautious footnote on ChatGPT — always good to keep the robots honest. But I’ve got to call out the link you dropped from the Law Society article like it was gospel.

🧪 Let's Talk About That 58% "Hallucination" Claim:​


The article you cited sounds alarm bells about GPT-4 hallucinating at a 58% rate — but that figure only makes sense if you're testing GPT like it's a Jeopardy contestant under pressure to regurgitate perfect legal citations on command.
  • The paper itself admits it evaluates GPT using zero-shot prompts, no context, and no fine-tuning — basically like asking your car to fly and then blaming it when it doesn’t.
  • It also doesn’t distinguish between critical errors and non-material ones, which in legal analysis is like equating a typo with perjury.
This is the equivalent of saying, “I asked my coffee machine to bake a cake and it failed. Therefore, coffee machines are unreliable.”

🧠 ChatGPT Is a Tool — Not a Wizard or a Demon

Nobody here (especially not @itsupportguy) is suggesting ChatGPT should be substituted for evidence in court. But dismissing its value based on “hallucination rate” is like saying law students shouldn’t be trusted because they sometimes cite outdated cases.

The real-world use of GPT in family law (or any law) is about:
  • Helping self-reps make sense of complex issues.
  • Structuring arguments more clearly.
  • Identifying possible themes or contradictions (yes, even in affidavits).
  • Drafting communications that don’t sound like they were written at 2AM on a stress high.

🧱 Re: The Article

The Law Society piece reads like it’s terrified that the profession might lose its gatekeeping badge. It wraps an alarmist stat in the cloak of academic caution — but forgets that actual legal reasoning requires nuance, not scare quotes. It’s not science; it’s CYA policy writing.

If we’re quoting from Oxford’s Journal of Legal Analysis to shut down LLMs in legal work, we may as well shut down legal Twitter, Reddit, and this very forum — where “hallucination” is sometimes a badge of honor.

👨‍⚖️ Final Verdict​


Yes, ChatGPT can be wrong. So can case law interpretations, section 30 assessments, judges, and — dare I say — lawyers who read one article too many.

Use your tools wisely. Don’t treat them like prophets. But don’t treat them like pariahs, either.
 
Our lawyer sent us this recent case that has been going around... Counsel used AI and submitted to the Court made up citations of cases that didn't exist. She was up for contempt...

That is a crappy lawyer who doesn't know how to use the tool. Its an expert system but, the expert driving it clearly didn't know how to use it. Sort of like blaming General Motors for making a car that was used as a weapon by someone. Or blaming DeWalt for the crappy job that someone did installing a kitchen. LOL. Or blaming MS Word for your failing grade in high school.

Lazy lawyers exist everywhere. Imagine what that lawyer would have done without the tool. When a "tool" uses a tool and doesn't do their job as required by their governing body. Expert systems guided the largest case this forum and the province of Ontario has ever seen. More case law than any of the anonymous lawyers are claiming against it. But, it was in the hands of an expert who "devastated" both the OCL and the other party's lawyer.

There are layers of governance control in AI platforms. If you use them like they are Google's "I'm feeling lucky" button on searching you may end up on a page that has nothing to do with what you were searching for. There are potentially 9 layers of governance in various AI platforms.

In ChatGTP PRO (200$+ Versions):

Tiers of Governance in ChatGPT

  1. Tier 0 – Core Alignment Kernel
    • Hardcoded ethical rules (e.g., no harm, no lies, no illegal content).
  2. Tier 1 – Moderation and Safety Filters
    • Real-time content screening for sensitive or prohibited material.
  3. Tier 2 – Behavioral Policy Layer
    • Enforces OpenAI’s rules on helpfulness, honesty, tone, and limitations.
  4. Tier 3 – Personalization and Memory
    • Adjusts responses based on your preferences, session history, or saved memory.
  5. Tier 4 – Output Formatting and Intent Matching
    • Shapes responses based on what you’re trying to do (e.g., code, legal doc, advice).
  6. Tier 5 – External Compliance and Deployment Rules
    • Follows laws, API limits, and platform-specific restrictions.
  7. Tier X – Ethical Overrides and Human Intervention (Shadow Layer)
    • Routes edge cases (e.g., threats, abuse reports) for manual review or suppression.

How to Keep AI Accountable​

  1. Ask for verbatim citations only. They will indicate when they are citing real law.
  2. Provide the case name or CanLII URL. They can analyze real text perfectly.
  3. Use the web tool (if enabled) and I can fetch real-time CanLII decisions

Common Triggers for Fabricated Case Law​

  1. User prompts to summarize a legal concept
    • AI might synthesize “plausible-sounding” precedent to illustrate a point (e.g., a made-up Ontario family case about shared custody). These can look real but don’t exist.
  2. User asked for a specific citation or example without grounding
    • If no real match is in my training data, AI may invent a name, year, or court level that seems legitimate based on linguistic patterns.
  3. Mixed jurisdictions or legal systems
    • AI might incorrectly attribute U.S. legal structure (e.g., Miranda rights) to a Canadian context or vice versa.
You need to setup your environment correctly. Go in raw and you will get raw results. The problem is that AI is not "intelligent" and the term AI is all wrong in industry. They are expert systems that often require someone with some level of knowledge to drive them. They don't read minds. You need to setup the environment correctly.

As well, 99% of users don't know how to use the project structure and create isolated regressive content stores in AI. That is because they get stuck on the "intelligence" component of the terminology incorrectly. Remember, many lawyers still use Word Perfect for drafting... So an advanced expert system may be out of most of their abilities. But, in 5-10 years... The next generation will level the playing field and the dinosaurs and luddite lawyers will be mostly retired or have angry clients when they are run over by highly educated lawyers who know how to use the tool and the fast advancements coming... Like 4o-Law... I have access to it.

The problem most users have is they are trapped in Layer 2 governance... They don't know how to get to Tier X in their instruction sets. Think of this way lawyers... You are using the tool at the same layer your most nightmare clients are using. They don't follow instructions, they can't articulate what their evidence is. They can't give instruction. You are left holding a bag of nonsense. AI platforms are more polite that humans and when given a bag of garbage they give the same garbage back.

There are instructions you can give to persist information and not be bound to a single session. But, you actually have to learn how to use the platform. Sort of like how you have to figure out how to use Document Management in Word (or Word Perfect if you are that kind of lawyer...) I can't tell you how many lawyers still use a file system to store critical information on their local PCs... Its sad that in 2025 that the majority of lawyers don't use technology correctly.
 
Follow-up clarification for future readers

Since this thread, the Ontario Superior Court of Justice has issued an updated Consolidated Provincial Practice Direction for Family Proceedings, which now expressly addresses the use of artificial intelligence in court materials.

The clarification in that Practice Direction aligns with what courts have long done in practice:

Courts do not inquire into how materials were produced.
They hold parties responsible for accuracy, verification, and responsiveness once defects are identified, regardless of the tool used.

The Practice Direction makes this explicit by placing accountability on the litigant or counsel using AI, emphasizing:
  • verification of authority,
  • correction of errors once identified,
  • and consequences only where misuse persists.
In other words, the judicial concern is not novelty or tooling.
It is persistence after correction.

I’ve written a separate post that steps back from technology entirely and explains how Ontario family courts assess credibility and reliability longitudinally, including why:
  • initial error is expected,
  • correction is neutral,
  • and repetition after guidance becomes evidence.
That framework applies equally to:
  • informal agreements,
  • lawyer-drafted materials,
  • expert reports,
  • and analytical tools, including AI.
For readers interested in judicial method rather than tools, the follow-up is here:

⚖️ Why Warnings About AI Miss the Judicial Question

This is not advice or advocacy. It is an explanation of observable judicial practice, now expressly reflected in the Court’s own Practice Direction.

Posted only to connect the threads, not to reopen debate.
 
Back
Top