kaal:claim:4796714-004
Ex-post governance, which applies regulation only after AI systems are developed and deployed or after large language models have been pretrained on existing proprietary datasets, fails to address risks and biases preemptively.
Source quote, verbatim
The traditional ex-post governance methods, where regulations are applied after AI systems are developed and deployed, or were pretrained LLMs models were trained on existing proprietary datasets, often fall short in preemptively addressing risks and biases.
From
Cite as
Holds when
Classification
failuresupport: arguedfailure: ex-post timing gapfamily: regulatory-lagai-and-agentsrisk-and-incentivesgovernance-design
Related claims
Verify