2026-08-19 — A concrete evidence gap behind the “40M ask ChatGPT” figure

• A STAT News piece documents a “shadow medical system” — 40M+ Americans ask ChatGPT health questions daily — alongside a companion JAMA Network Open study by the same authors finding AI models fail to produce a comprehensive differential diagnosis in >80% of cases when given only incomplete initial patient info, despite doing well on complete case data. A concrete evidence-gap data point for this thread’s standing “is this push real, and is it safe” question — extends rather than resets the thesis. (STAT News)

2026-08-15 — Google ships its health push into hardware, but the MH layer doesn’t move with it

• At its “Made by Google” event on August 12, Google unveiled the Pixel Watch 5 with passive blood-pressure and insulin-resistance tracking plus a deeper Gemini integration, and a $9.99/month Google Health Premium coaching tier — a second giant, after OpenAI’s Sheba Medical Center deal, moving its health push from software announcement into a shipping consumer product with real distribution. This is the “real vs. announced” test this meta-thread tracks landing in Google’s column, but the mental-health layer is the same carryover stress/sleep tracking from the prior generation — no new MH-specific feature accompanied the launch, unlike Google’s April Gemini crisis-hotline update, which this thread had not yet logged. (TechCrunch)

2026-08-05 — A cross-actor data point: Gen Z/millennials now default to AI before a doctor

• Aflac’s fourth annual Wellness Matters survey found 76% of Gen Z and 63% of millennials turn to AI first for health guidance, ahead of seeing a professional — versus just 3% of Gen X and 2% of boomers; 18% of Gen Z and 14% of millennials delay booking an appointment until after exhausting AI/digital tools. This is the vendor-neutral, demand-side evidence this meta-thread has been missing — the behavioral shift pulling every giant into health, not any one company’s product move. Verified: Aflac’s own survey, corroborated by the article’s cited figures (survey publish date 2026-08-05). (Business Journal Daily)
• The mental-health carve-out inside the same survey: despite skipping primary care (only 48% of Gen Z has a regular doctor), Gen Z is comparatively more willing to see a mental-health therapist (25% have engaged one) than older cohorts — the AI-first substitution effect is uneven across care types, cutting deeper into primary/preventive care than into therapy specifically. (Business Journal Daily)

2026-07-28/29 — OpenAI’s health push goes institutional; a HIPAA gap flagged

• Sheba Medical Center (Israel) partnered with OpenAI on what multiple outlets independently headline as the first international ChatGPT-for-healthcare deployment — the first global, hospital-wide rollout of ChatGPT for Healthcare, giving physicians, nurses, researchers and operational staff access to models that synthesize peer-reviewed literature, clinical guidelines and internal protocols with verifiable citations. Moves the 07-23 GA consumer launch into actual institutional/enterprise healthcare delivery — the concrete “is this push real” answer this thread has been watching for. Confirmed via direct fetch of HIT Consultant + corroborated by MobiHealthNews and a press release. (HIT Consultant)
• A single, lower-reliability outlet claims ChatGPT Health carries no HIPAA coverage as of 2026 — consistent with this thread’s own evidence-gap framing (a health layer that walls off liability rather than becoming a regulated clinical product), but single-sourced — flag, not confirmed.
• One low-confidence report of a second entrant — Perplexity launching a health platform with Apple Health integration — single non-major outlet, worth a follow-up crawl before adding a sixth entity to this meta-thread’s tracked list.

2026-07-23 — First real entry: OpenAI’s health layer goes live, MH still absent

• OpenAI’s ChatGPT Health went GA to all US users 18+ (07-23), linking Apple Health data — the same week as a wave of suicide/ psychosis-reinforcement suits and a 42-state AG probe of the general model. Reads as the evidence-gap thesis in miniature: a frontier lab building a formal “health” layer precisely because unmanaged de-facto health use (>1M/week suicidal-planning signals on the general model) is the liability. MH specifically is not a named feature of the launch — it walls off general physical-health tracking while the companion-chatbot harm keeps accruing on the same underlying model. Cross-refs OpenAI Health and The Reckoning (where this was first logged 07-24 but not yet folded into this meta-thread). (TechCrunch)

2026-07-24 — Opened

• Thread opened (ben-steer) — meta-thread — groups the child threads below.

← Backstory