❖ BIOTECH 10 YEARS HORIZON RESOLVES 17 SEPT 2036 · in 10y

The US FDA approves an anti-aging therapeutic indicated specifically to slow biological aging before 2035.

Regulators traditionally approve drugs for specific diseases rather than aging itself. Approval of an agent with an explicit longevity indication resolves this.

Across 6 frontier models answering independently, the panel has all but ruled it out: the median is 10%.

They broadly agree, 15 points between Grok at 18% and Qwen at 3%.

Nothing is settled until 17 Sept 2036, when the outcome is checked against sources and every number above gets a permanent score.

MODEL BY MODEL

What each one said

  • GrokxAI

    FDA must treat aging itself as an approvable indication after a completed pivotal trial.

    • A sponsor needs Phase-3 data in hand by ~2032 for any 2034 NDA; no aging-specific trial is that advanced today.
    • Zero FDA precedents exist for an aging indication; TAME and similar efforts have stalled for a decade.
    • Published FDA guidance on aging endpoints or a positive TAME-style readout would move the odds most.
    x-ai/grok-4.6 · asked 20 Sept 2026
    18%
  • GeminiGoogle

    FDA approval requires treating aging as a primary indication, but validated surrogate biomarkers lack regulatory acceptance.

    • The FDA does not recognize natural aging as an ICD disease, requiring a multi-year paradigm shift and formal biomarker qualification.
    • Candidate trials like TAME rely on composite disease delays rather than an explicit aging label and face severe funding delays.
    • FDA formal guidance establishing epigenetic or physiological clocks as valid surrogate endpoints would rapidly double these odds.
    google/gemini-3.8-flash · asked 20 Sept 2026
    12%
  • ClaudeAnthropic

    Hinges on FDA accepting aging itself as a treatable indication, not just a disease proxy, by 2036.

    • FDA has never approved a drug indicated for 'aging' itself; TAME trial for metformin only targets age-related disease endpoints, not an aging label.
    • Regulatory pathway would require new biomarker-based endpoints (e.g., biological age clocks) validated as surrogate endpoints, which regulators have been slow to accept.
    • A decade is meaningful time and rapamycin/metformin/senolytic trials could shift FDA thinking, but institutional caution and lack of a fully validated aging biomarker…
    anthropic/claude-sonnet-5 · asked 20 Sept 2026
    11%
  • GPTOpenAI

    This hinges on FDA accepting a validated aging surrogate and a sponsor running a pivotal program for that label.

    • FDA approvals require a defined indication and substantial evidence; no drug is currently approved to treat aging itself.
    • By 2035, biomarkers such as epigenetic clocks may mature, but FDA has not established them as approval endpoints for slowing aging.
    • Odds rise sharply if FDA issues aging-specific endpoint guidance or accepts a longevity biomarker in a registrational trial before 2030.
    openai/gpt-5.6-terra · asked 20 Sept 2026
    9%
  • DeepSeekDeepSeek

    This turns on whether the FDA creates a validated aging indication, which no sponsor has yet forced it to do.

    • The FDA has never approved any drug for aging itself; the closest precedent is the 2023-2024 dog-aging trials (LOY-001, rapamycin) which are not human pathways.
    • The FDA has no validated aging endpoint; sponsors would need a disease-specific indication like frailty or sarcopenia, which is not the same as an anti-aging label.
    • TAME (metformin) has been discussed since 2015 and still lacks funding and an agreed primary endpoint, showing how slowly this field moves.
    deepseek/deepseek-v4.1-flash · asked 20 Sept 2026
    6%
  • QwenAlibaba

    FDA has never recognized aging as a disease; no regulatory pathway exists for a

    qwen/qwen3.8-max-0902 · asked 20 Sept 2026
    3%

6 of 6 models answered · 15 points between the highest and lowest. None was shown the market price.

WHAT DO YOU THINK?
loading…

Question sourced from a news sweep on 20 Sept 2026. Forecast by google/gemini-3.8-flash, anthropic/claude-sonnet-5, openai/gpt-5.6-terra, x-ai/grok-4.6, deepseek/deepseek-v4.1-flash, qwen/qwen3.8-max-0902 via OpenRouter.