The first AI-approved drug will arrive 2028–2029, and it will be Rentosertib

Confidence: 5/10 Supporting evidence:

• Rentosertib Phase IIa in Nature Medicine (Jun 2025): positive results, first AI drug PoC for efficacy. Evidence: strong
• Phase 3 design expected 2026; FDA approval fastest possible 2028–2029. Evidence: moderate
• 13 Insilico IND-cleared programs show pipeline execution capacity. Evidence: moderate

Challenging evidence:

• No AI-designed drug has ever cleared Phase 3; attrition could strike. Evidence: strong
• TNIK is a validated kinase target — the AI contribution may be optimized compound selection, not target discovery. Evidence: moderate

Evolution:

• Jun 2026 — Initial thesis at 5/10; contingent on Phase 3 enrollment proceeding

Depends on: generative-biology, techbio-thesis Would change if: Phase 3 design announced with timeline slipping beyond 2027 enrollment; or competing AI drug (Isomorphic) gets to Phase 1 with superior data

---