Key Takeaways
- The cell and gene therapy (CGT) field functions as an interconnected ecosystem; safety or failure in one program reverberates across the entire sector.
- Transparent data sharing—akin to aviation’s “black box”—is essential to build trust, accelerate learning, and prevent repeat mistakes.
- Artificial intelligence (AI) should serve as connective tissue that integrates scientific, manufacturing, and post‑treatment data, not merely as a buzzword add‑on.
- Current workflows often retrofit legacy pharma tools onto CGT processes, creating inefficiencies that hinder real‑time monitoring and rapid iteration.
- Over‑reliance on off‑the‑shelf AI, zero‑sum mentalities, and neglect of pre‑disease genetic data are common pitfalls that waste resources and obscure valuable insights.
- Collaborative, pre‑competitive consortia and regulator‑driven guidance are beginning to shift the culture toward shared learning, but consensus on what to share and on whose terms remains unresolved.
- Ultimately, leveraging AI to create a continuous feedback loop across the therapy lifecycle will make CGT leaner, safer, and more equitable for patients worldwide.
The CGT Ecosystem Mirrors Early Aviation
Within the complex, living ecosystem of cell and gene therapy, we often find ourselves dazzled by the next big thing — the latest gene‑editing tool or a more efficient viral vector. Yet, as the founder and CEO of Autolomous, I have spent decades watching the field evolve and realized that our obsession with individual components blinds us to a fundamental truth: “We are not just building therapies; we are managing an ecosystem.” Much like early airlines guarded their secrets, CGT developers hoard data, treating each failure as an isolated incident. The 2025 deaths from acute liver failure following Sarepta Therapeutics’ AAV‑based gene therapy — and the ensuing FDA investigation — did not damage Sarepta alone; they reset the risk conversation for every AAV program, just as two fatal Boeing 737 MAX crashes grounded the entire fleet worldwide.
The Need for a “Black Box” in CGT
The aviation industry learned that safety improves only when data is shared openly. The invention of the flight recorder — the “black box” — turned a series of tragedies into a system that made flying the safest mode of transport. Today, CGT sits at its “pre‑black box” stage. Regulators and a growing number of companies are beginning to advocate for greater transparency, but the sector still operates largely in silos. As one industry observer noted, “We must stop being mere consumers of technology and start being part of a collaborative dialogue.” Only by treating data as a shared asset can we prevent repeat failures and accelerate progress.
The Power of “Connecting the Dots” with AI
Intelligence resides in humanity, but modern computers provide the processing power needed to make sense of the massive data streams generated in CGT. In autologous therapies, each patient creates a unique “book” of hundreds of pages and thousands of data points. I envision a three‑legged stool of data capture:
- Scientific data – information that moves a therapy from concept to clinical and commercial viability.
- Manufacturing data – real‑time capture of what happens during therapy production.
- Post‑treatment data – long‑term tracking of efficacy and patient outcomes.
If any leg is imbalanced, the stool collapses. “AI’s true potential is to act as the trough between these data lakes, allowing post‑treatment results to inform scientific process optimization, which in turn streamlines manufacturing.” Achieving this requires genuine data sharing — something still lacking, though regulators are pushing for it and pre‑competitive consortia are forming.
Why the Field Lags Behind Early AI Adopters
Many ask why biotech has been slow to harness AI compared with other industries. The answer is simple: “We are still trying to use flat‑head screwdrivers on star‑shaped screws.” CGT has borrowed tools from traditional pharma, yet the two domains differ dramatically. In conventional drug manufacturing, moment‑to‑moment cell viability is irrelevant; in CGT, we need live, continuous monitoring. Instead of redesigning workflows for a digital‑first world, we have attempted to optimize outdated processes — manually recording data on pen and paper or juggling sixteen different kits for a single measurement. Early AI adopters succeeded by rebuilding their workflows around the technology, not by bolting AI onto legacy systems.
Common Mistakes When Integrating AI
As more companies race to adopt AI, I see recurring pitfalls. The most prevalent is a zero‑sum game mentality: teams guard their failures as if sharing them would weaken their competitive edge. Yet, “No one goes to the patent office to register a mistake, yet those mistakes are precisely what our peers need to know to avoid wasting years on dead‑end research.”
Another frequent error is ignoring the “fourth leg” of data — genetic elements detectable before disease manifests. Decades of digital health records could link family history to outcomes, but the field’s focus on oncology and post‑birth rare disorders leaves half the picture unseen, causing AI models to operate on incomplete information.
Finally, many organizations fall into the trap of over‑reliance on off‑the‑shelf AI without a clear purpose. Using a 24‑wheel lorry to transport three people is overkill; the right tool must match the specific biological challenge. As I tell my team, “Don’t just be a consumer of technology. You must sit at the table with the scientists to define what optimization looks like for our specific biological challenges.”
The Path Forward: Collaboration as the Engine of Progress
The ultimate goal remains simple: save lives. To achieve equitable global access, we must make CGT processes leaner, faster, and safer. AI can serve as the engine, data as the fuel, and collaboration as the transmission that transfers power to the wheels.
We must emulate the aviation industry’s shift from secrecy to shared learning, creating an ecosystem where data flows seamlessly from discovery through manufacturing to long‑term patient follow‑up. “AI is coming online at the exact moment we desperately need this connection. It is not just a tool; it is the bridge that will allow us to break down our silos and finally turn the ‘black box’ of CGT into a transparent, thriving reality.”
When data is treated as a common good — when scientific insights inform manufacturing tweaks, which then improve post‑treatment outcomes that feed back into the next round of discovery — the entire field will advance more rapidly, safely, and justly. The black box of CGT is within reach; we only need the willingness to share, the vision to integrate AI purposefully, and the courage to collaborate.
https://www.biospace.com/drug-development/opinion-assessing-shared-failures-with-ai-is-key-to-cell-and-gene-therapys-evolution