An enormous limitation of ESG disclosures is the partitioning of the metrics into silos, where the interdependence of the numbers is not factored in, how do two metrics intertwine in a particular context to produce risk in a real time manner which is useful to the decision maker? The sociopolitical context can attenuate or amplify the risk, which is missing from the ESG discourse as standardization for comparability is the impulse rather than how is the data valuable for investor decision making at this moment.

AI Models are great, but real risk is experienced on the ground where the incident happens, prior to ossifying into an incident statistic tucked into the reams of the disclosure deluge.