Identity Stability as a Missing Risk Surface in Programmable Biology

Biological interventions are increasingly durable, adaptive, and state altering. Genome editing, cellular reprogramming, regenerative strategies, and living cell therapies persist over time and interact continuously with host systems rather than producing transient or terminal effects. Existing safety and evaluation frameworks were developed for interventions whose risks were primarily acute, local, or short-lived. As a result, a distinct class of risk has emerged that is not systematically assessed prior to biological exposure: the risk that an intervention fails to preserve its intended biological identity over time under real biological pressure. This working paper introduces Identity Stability Risk Assessment (ISRA), a pre-in vivo conceptual framework for reasoning about identity durability in programmable biological systems. ISRA does not predict therapeutic efficacy, nor does it replace empirical safety testing. Instead, it defines identity stability as an explicit risk surface and provides a structured way to evaluate where and how biological identity may drift, fragment, or reorganize after irreversible commitments have been made. The paper formalizes the concept of identity stability, explains why identity-related failure modes tend to emerge late under current evaluation paradigms, and outlines how ISRA enables earlier, regulator-aligned reasoning about long-term biological behavior. An illustrative appendix demonstrates how the framework may be operationalized as a structured risk profile, without presenting new experimental data. This manuscript is intended as a foundational systems-level contribution to discussions of risk, durability, and governance in long-lived biological interventions.