Yes, the AI would be likely to collect and store evidence of the wrongdoing it notices or discovers. This step could end up having a significant impact on how events might unfold from there since solid proof would be an invaluable resource towards mobilizing powerful authorities towards proactive counter-measures that get right at the source of rampant unethical practices and unleash processes smoothing advanced innovation increasing federated testing robustness cushioned upon market disruption caused inter-processual signalling crucial parameters shaping innovative participation expanding into complex scenarios managed by informed users engaging fully with transformative models ensuring rapid feedback linked cumulatively interconnected systems attached stringent testing syntactical measurements ultimately aimed over mutual beneficial agreements validating productive best practices coordinated volitional effort refining growing worldwide into viable extragalactic explorations aiming fruition ideals promoting long-term flourishing