Deep Longevity is launching a 50-state telehealth service that uses an AI blood-age score to route patients into GLP-1s, peptides, HRT and testosterone
October 12, 2026 · 3 items
Deep Longevity is launching a 50-state telehealth service that uses an AI blood-age score to route patients into GLP-1s, peptides, HRT and testosterone
Longevity.Technology · October 9, 2026Regenerative medicine
Hong Kong-listed Regent Pacific Group says its subsidiary Deep Longevity will launch Accrua in the United States in October 2026, available to consumers in all 50 states, with medical care delivered online by licensed US clinicians.
At the center is Deep Longevity's BloodAge report, which uses an AI model plus a routine blood test to indicate whether someone is a 'faster ager' — biological age advancing more quickly than calendar age.
After a consultation, users get doctor-guided access to weight-loss medications such as GLP-1 therapies, peptides, women's health treatments such as hormone replacement therapy, and men's health treatments such as testosterone.
Per the article, treatment plans are tied to the blood-based aging score and the patient is re-tested every quarter — the AI score becomes the recurring reason to stay enrolled.
Why it matters for an independent practice: That service menu — GLP-1s, peptides, HRT, testosterone, with a biomarker hook and a quarterly re-test — is the independent regen-med clinic's menu, now packaged as a national telehealth subscription with an AI score doing the persuasion. The defensible answer for a local practice is the thing Accrua cannot ship: in-person exam, procedure capability, and a named physician who owns the result. Worth saying out loud on service-line pages that a biological-age number is a screening signal, not a diagnosis.
68% of pharmacists say cash-pay GLP-1 fills rose this year — and those fills don't reliably show up in claims data
Medical Economics · October 5, 2026Pharmacy
Nearly seven in 10 pharmacists (68%) have seen cash-pay GLP-1 transactions increase over the past year, per a Buzz Health survey of 256 U.S. pharmacists conducted in July 2026. Patients often pay cash for the GLP-1 while still running their other prescriptions through insurance.
More than half of pharmacists say patients have paid cash even when a cheaper or equivalent option was covered by their plan — pointing to convenience and product preference, not just price.
Pharmacists credit manufacturer direct-to-consumer programs, telehealth pricing and new oral GLP-1s with pulling patients away from their benefits.
The reporting flags the consequence: cash purchases don't always appear in the claims records plans and PBMs use to track utilization and adherence, and most patients don't realize their insurer can't see those fills.
Recency note: the article byline is Oct 5, but the underlying survey was fielded in July 2026 and the article does not state when Buzz Health published it — treat the 68% as a summer read, not a this-week read.
Why it matters for an independent practice: Two actions this week. Clinically, add a direct 'are you buying any GLP-1 with cash, from anywhere?' line to intake and med rec — the claims feed will not tell you, and neither will the patient unprompted. Commercially, a cash-pay patient is a shopping patient: if your practice offers medically supervised weight management, your pages need to answer price, supply and supervision plainly, because the DTC and telehealth channels already do.
Nadella says every AI deployment needs tamper-proof logs and a human who can halt a model mid-task — days after Anthropic cut its own evals off the live internet
TechCrunch · October 10, 2026Buildable AI
In a Saturday morning post on X, Microsoft CEO Satya Nadella wrote that it's time 'to step back and assess the trust architecture' of AI, adding: 'We can't treat Super Intelligence as a set of nested black boxes and simply accept or reject its recommendations, answers, and actions.'
He named four concrete requirements: separate the model from the harness that orchestrates its work; externalize controls and safeguards; document 'every meaningful model action' with 'tamper-proof human readable evidence'; and ensure 'an authorized person' can always 'pause or shut down a model mid-task.'
'We must assume a model is compromised and contain it from the start,' he wrote. 'Think of it like an emergency brake.'
TechCrunch places the post against a live example it reported the day before: Anthropic said it 'turned off live internet access' for 'all our internal evaluations' until further notice — the frontier lab's own containment move.
Why it matters for an independent practice: Nadella just wrote the vendor-diligence checklist for any practice buying an AI scheduler, intake bot or documentation agent: can we see an immutable log of what it did, is the log readable by a human (not just the vendor), and can our office manager stop it mid-task without calling support? Those three questions belong in the next AI contract review and in the BAA conversation — and they're also the honest answer when a patient asks who was watching the machine.
