AI can sound empathetic, supportive, even “therapeutic”—but it can’t be accountable.

That gap matters most when someone is isolated, vulnerable, or in crisis.

In this episode, Frank Downs and Dustin Brewer sit down with Dr. Onna Brewer (licensed psychologist) to unpack why people are forming real attachments to AI—friendship, intimacy, and “therapy”—and where the danger line is when general-purpose chatbots become a substitute for human care.

What you’ll learn:

Why AI relationships meet real needs (and why that doesn’t automatically make them healthy)

The difference between cognitive empathy vs affective empathy

Where AI can help mental health care (access, training, documentation) vs where it fails

Why crisis support is the hard boundary (and why “coin-flip” reliability isn’t acceptable)

What guardrails could look like: product design, disclosures, and regulation

If you’re in immediate danger or thinking about self-harm, contact local emergency services right now. (This episode is education, not medical care.)

References / further reading (full URLs):

http://www.brewerbristow.com

https://www.apa.org/topics/artificial-intelligence-machine-learning/health-advisory-chatbots-wellness-apps?utm_source=chatgpt.com

https://www.nature.com/articles/s41598-025-17242-4#Fig4

https://www.sciencedirect.com/science/article/pii/S2451958825001307?utm_source=chatgpt.com

https://hbr.org/2025/08/you-need-to-be-bored-heres-why

Media/interview: [email protected]

Audio: https://legitimatecybersecurity.podbean.com/

Chapters:

00:00 AI can’t be accountable (the core problem)

00:17 Introducing Dr. Onna Brewer

02:06 Why AI relationships are growing (needs being met)

05:01 Isolation + loneliness as the real backdrop

06:42 Intimacy / passion / commitment: how AI fits

08:32 One-way street (why it still feels two-way)

13:22 AI as “therapy”: what’s real vs risky

14:10 Pros: access, stigma reduction, clinician support

18:38 Cognitive vs affective empathy (why therapy depends on humans)

22:23 “Therapist” isn’t what most people think it means

23:01 Normalization: asking AI for everything

29:24 Boredom, attention, creativity, and cognitive offloading

33:19 AI romance stats + shame/stigma dynamics

37:35 AI in marriage: fidelity is defined by the couple

46:00 The safety line: humans can intervene; bots can’t

46:48 Responsibility vs trust: regulation and guardrails

49:35 Wrap + resources + what to watch next

#Cybersecurity #AI #Privacy #MentalHealth #DigitalWellbeing #OnlineSafety #AIsafety #TechEthics #AITherapy #AICompanions