ChatGPT has made several announcements recently, including its new Teen Safety Blueprint. This is a framework aimed at giving minors a safer, age-tailored AI experience, and was likely created as a response to mounting criticisms and lawsuits by putting down a marker for how it thinks teen use should look.

Here are a few takeaways:

Different experiences by age. The Blueprint states that ChatGPT’s responses to a 15-year-old “should differ from the way it responds to an adult.” It also describes using age-estimation tools to identify minors based on factors like usage patterns and system data, not personal details, to deliver age-appropriate safeguards.

The company hasn’t yet explained how the model will respond differently to teens, only that interactions will be filtered through age-appropriate safeguards and avoid serving as a ‘substitute therapist or best friend.’”

Default to the safer side. If the system can’t determine someone’s age, it defaults to treating the user as under 18, so the safer pathway is chosen automatically.

Guidance for families. New parental controls allow caregivers to link accounts, set blackout hours, and adjust privacy settings, offering a way to support teen AI use rather than simply restrict it.

Why This Matters for Therapists

“Stay in the boat” is a phrase I use with families at the beginning of FT-AN/FBT treatment. No matter how stormy the sea, no matter how wild the weather gets, we stay in the boat. We don’t jump out at the first wave, we don’t swim off in panic, we don’t let the conditions convince us we’re sinking. We stay. We hold the line. We let the storm pass, because it will. I say it in relation to boundaries, too.

This week, I have needed my own advice. We had to make the decision to put our dog, our fur daughter, to sleep. We have lived this before, earlier this year and it’s still a decision that feels impossible even when unavoidable. Those noisy heartbreaking decisions, where love demands something of you that courage has to carry. I kept thinking about the boat, about staying with the discomfort, the grief, the conviction that we were doing the kindest, hardest thing. And yet wanting to climb out of the boat entirely. Frankly wanting to rip my skin off. And knowing I couldn’t.

Professionally and personally, decisions such as these sometimes feel like standing on a cliff edge with no map. The ethically heavy ones, never come with certainty. To me, staying in the boat isn’t about being calm or brave or certain; it’s about staying present when every part of you wants to burn it down. The stubborn decision to remain with the pain instead of abandoning it.

Therapists can use this episode as a reflective viewing assignment with clients exploring family roles, cultural identity, codependency, and the impact of addiction. It’s a vivid illustration of what happens when unspoken pain meets holiday rituals, and how affection, loyalty, and dysfunction often coexist in the same room.

Encourage clients to notice:

“Fishes” opens a conversation about how our families teach us to love, protect, and sometimes disappear ourselves in the process. It’s messy, loud, and deeply human, just like most of the systems we come from.

When Marianne and I had our weekend retreat in Cornwall a week and a half ago, I could not get over its beauty. Even looking at the same view of the sea, with abundant whitecaps, an endless beach at low tide, and dramatic cliffs, never got old. One of my secret coping skills is diving into Jenny Colgan novels, and being in Cornwall gave me all the cozy seaside-bakery-meets-new-chapter energy her books are known for. Colgan writes stories of young Brits who escape to coastal towns to start over, finding purpose, love, and sweetness in the everyday. Standing on those windswept cliffs, I feel like I stepped into one of her novels.

I clocked off to make space for it, and I’m grateful to have a boss who understands that losing an animal you’ve loved for 14.5 years isn’t “less” grief than any other, and who treats me with the same compassion they’d offer someone with human children. That matters.

So we’ve taken the time we needed. To feel it properly. Time to hold on to the moments we had. Time to bury her the way we wanted to. Time to help her furry sisters make sense of their missing leader. And time to be with my Mum and Dad, who truly are her furry Grandma and Grampy, as they navigate their own heartbreak.

November 20, 2025 | 1–2 p.m. ET | Recording available

Heard and Blueprint are co-hosting a live, therapist-specific tax prep workshop. It’s perfect timing before the holiday rush starts.

In one hour, presenters Andrew Riesen, CPA & Forrest Perry, CPA from Heard will cover:

If taxes usually make your eyes glaze over, you’re not alone. Hope to see you there!

Register here.

Note: Ann works as a contractor with Blueprint; this mention is included independently and without compensation.

A new observational study on Ash, Slingshot AI’s mental-health companion, has been circulating on LinkedIn this week — accompanied by glossy headlines about reduced loneliness, increased social support, and users feeling more hopeful and socially engaged.

On the surface, it sounds like the holy grail: an AI system that not only talks to you but sends you back into your real life more connected than before.

According to the reported findings:

If true — actually true — this would be a breakthrough.

But that’s the problem: we don’t yet know if any of this is true.

What matters most about this story isn’t the results. It’s how they were achieved and the methodology behind them.

On the viral LinkedIn post, psychologist and researcher Albert Wong published a surgical response, one worth every therapist’s attention. His central argument: the study’s scientific framing is so weak and so misrepresented that the results are far less meaningful than the hype implies.

Why the Study Isn’t What It Claims to Be

1. It’s not a clinical trial. It’s a self-selected customer sample.

The authors reference CONSORT standards — guidelines used for randomised controlled trials.

But Ash didn’t run a trial.

It ran an observational cohort study using only people who stuck around long enough to provide data.

In research terms, that’s survivorship bias disguised as evidence.

The appropriate standard wasn’t CONSORT at all.

It was STROBE — the guideline for observational studies.

Referencing the wrong guideline makes the work look more rigorous than it is.

2. All the users who dropped out simply disappeared.

As Wong put it, this is like a restaurant announcing “our food is clinically proven delicious” after surveying only the customers who came back multiple times and ignoring everyone who tried the food once and never returned.

When your sample excludes people who have stopped using the product, you are effectively reporting only on fans.

3. No disclosure about payment or incentives.

The study never states whether users were compensated, how much, or how compensation might have influenced retention or mood scores. For any behavioural or engagement-based outcome, this is a basic transparency requirement.

When missing, it raises immediate red flags.

4. The danger isn’t that the findings are positive. The danger is that they will be cited as evidence.

This is where the real risk lies. Not in Ash’s existence. Not in users finding some comfort. But in companies and policymakers pointing to methodologically sloppy research as if it is high-quality, peer-reviewed proof of efficacy.

We’ve seen what happens when mental-health technologies grow faster than the science around them. We’re still living with the consequences.

Therapist Takeaways

AI in mental health will only be as ethical as the research that supports it. We cannot build a future of AI-enhanced care on selective samples, inflated claims, or misapplied methodologies.

And as clinicians, this is our role: to be excited, yes — curious, hopeful, open — but also skeptical and to insist on standards that match the stakes.

AI may well become part of the mental-health landscape, but if the foundation is shaky, everything built on top will eventually collapse.

Here’s how the votes shook out:

Apparently, our community runs on curiosity, ambition, and zero chill.

Comments from the Community:

“I'd love to travel the world and collect CE's while doing it, but not at an astronomical cost. ”