# AI longevity — five more years source review
Reviewed 2026-09-20. Original source: https://youtu.be/dQYKcjvXhIY?si=ktXHnbbeaBx8OUZn
Video: Please Stay Alive For 5 More Years
Channel: There's An AI For That; upload date 20260824; runtime 1206 seconds.

## Capture and duplicate handling
Metadata and 563 English caption segments retrieved. Full transcript and downloaded source extracts remain in the private research archive; public article uses original paraphrase and primary links. No full video or full captions rehosted. Captions may contain spelling/numeric errors and are not treated as authoritative wording for financial figures. Existing ER-100, OpenAI/Retro and Yamanaka coverage was checked; this source video had no exact URL/ID match. This is a distinct cross-field evidence review with related-article links, not a replacement of those older pages.

## Scope and classifications
The evidence ledger covers 13 major claim groups. It separates preclinical mouse/cell research, corporate laboratory results, trial registration/first dosing, disease-specific approval, clinical endpoints and forecasts. It is not an exhaustive audit of all biographical anecdotes, financial totals or first-in-history claims.

ER-100: sponsor's June 9 first-dose announcement does not give the actual administration date. ClinicalTrials.gov API was read directly; last update May 19, 2026, recruiting status posted, Phase1, open-label/non-randomized, estimated enrollment18, no results posted. No completed-cohort or human-efficacy claim. Registry safety descriptions should not be transformed into instructions for self-treatment.
OpenAI/Retro: >50-fold marker expression in vitro and other reported lab characterization, not human age reversal. Corporate/internal replication is not independent clinical proof.
Rentosertib: primary safety endpoint; secondary FVC signal; withdrawals and liver toxicity acknowledged. Full publisher HTML read directly after extractor gave incomplete content.
Casgevy: historical FDA December2023 disease-specific approval. No claim about current Canadian access or complete modern label scope.
Xenotransplant: NYU Looney update and newer MGB/Lancet Andrews case read. Graft failure, recipient death and later human transplantation are different outcomes. Lancet abstract and institution account available; full subscription paper not obtained.
MASAI: 2025 and 2026 reports are required for later claims; 2023 citation alone insufficient. Interval-cancer ratio CI crosses1; non-inferiority is not superiority or mortality benefit.
UNITY: source description incorrectly links a different diabetic-kidney senolytic paper. Correct UBX0101 sponsor release retrieved from GlobeNewswire after legacy IR domain was unhelpful. Not an independent news report.
Isomorphic timeline: specific January2026 cancer-drug clearance not corroborated. Reuters timeline and separate Insilico ISM8969 Parkinson's IND announcement distinguish companies and indications. Video does not name a drug; no definitive attribution of its intended source.

## Access and reporting limits
Some direct PubMed requests returned cookie challenges and some extraction requests timed out. Previously captured primary PubMed records/abstracts from this research run, cached paper text and accessible publisher/registry pages supplied the relevant results. NAD review interpretation is based on its accessible abstract; not a claim to have obtained the full subscription text. No paywalls bypassed, no trial enrollment or outreach undertaken. Exact source URLs are in source-urls.txt.

## Editorial conclusion
Real research advances do not establish a five-year guarantee, whole-body human rejuvenation or immortality. Maintain preventive care, screening and treatment. No dosing protocol, supplement sales or promise of access. Educational material, not medical advice.