Using authentication challenges to automatically obtain training data to train a machine learning model

A system comprising:

• a memory;
• and
• a processor communicatively coupled to the memory, the processor to perform operations comprising:
• identifying at least one generative machine learning model (MLM) to be trained using training data reflecting analytical responses of humans;
• and
• automatically collecting the training data for the at least one generative MLM from a plurality of users, wherein automatically collecting the training data comprises:
• providing an authentication challenge for each user attempting to access a resource, wherein the authentication challenge requests that a respective user of the plurality of users provide a set of responses comprising a first user response to a first sample, wherein the first user response indicates that the respective user is a human by identifying one or more errors in the first sample, and a second user response to a second sample, that the second user response indicates an analytical response of the respective user;
• and
• responsive to determining, based on the set of responses, that the respective user is a human, using the second user response provided by the respective user as part of the training data for the generative MLM.

A non-transitory computer readable storage medium comprising instructions for a server that, when executed by a processing device, cause the processing device to perform operations comprising:

• identifying at least one generative machine learning model (MLM) to be training using training data reflecting analytical responses of humans;
• and
• automatically collecting the training data for the at least one generative MLM from a plurality of users, wherein automatically collecting the training data comprises:
• providing an authentication challenge for each user attempting to access a resource, wherein the authentication challenge requests that a respective user of the plurality of users the provide a set of responses comprising a first user response to a first sample, wherein the first user response indicates that the respective user is a human by identifying one or more errors in the first sample, and a second user response to a second sample, wherein the second user response indicates an analytical response of the respective user;
• and
• responsive to determining, based on the set of responses, that the respective user is a human, using the second user response provided by the respective user as part of the training data for the generative MLM.