Read six research-led essays on evidence for competitive barriers in agentic and physical AI.
/ MoatDefender
Early open-source alternatives to Adobe and Office show why feature replication deserves a fresh test. Production quality remains unproven; defensibility increasingly needs evidence beyond code.
Read essay/ MoatDefender
The 7 Powers remain a useful base. AI changes the cost of imitation, so each claimed advantage needs a fresh test of its benefit, barrier and evidence.
Read essay/ MoatDefender
Agents may reduce migration work. Durable switching-cost claims need measured residual effort, customer benefit and evidence beyond contract friction.
Read essay/ MoatDefender
Calendar time can matter when advantage depends on accumulated operating evidence and constrained physical access. Elapsed time alone is no moat.
Read essay/ MoatDefender
Trust can become a scarce competitive asset when buyers can verify its basis. Security controls, provenance and certificates alone do not prove a moat.
Read essay/ MoatDefender
Measure the improvement a learning system produces, the cost of reproducing it and who has rights to the feedback. Counted records are only an input.
Read essay/ MoatDefender
The current reading assesses defensibility from evidence using a versioned draft method. Global rankings and expanded protection features from our roadmap are coming.
Read essay