Multi-level text and typing data authentication

A non-transitory computer-readable medium having program instructions stored therein that are executable by a computer system to perform operations comprising:

• detecting a set of input data at a computing device, wherein the set of input data includes text data and typing data;
• generating, using a first machine learning (ML) model, a first prediction for the set of input data, wherein the first ML model is trained to increase a similarity between embeddings of authentic text data and corresponding sequences of typing data included in the set of input data;
• generating, using a second, different ML model, a second prediction for the set of input data, wherein the second prediction is a feature vector output by the second ML model based on predicting an intended sequence of text based on the text data and a sequence of typing actions included in the set of input data;
• determining, using a classifier model, whether the set of input data detected at the computing device is authentic input, wherein the determining is performed based on the first prediction and the second prediction;
• and
• transmitting, to the computing device, an authentication decision, wherein the authentication decision is generated based on determining whether the set of input data is authentic.

A system comprising:

• a processor;
• and
• a non-transitory computer-readable medium having stored thereon instructions that are executable by the processor to cause the system to perform operations comprising:
• detecting a set of input data at a computing device, wherein the set of input data includes text data and typing data;
• generating, using a string-level model, a string-level prediction for the set of input data, wherein the string-level model is trained to increase a similarity between embeddings of authentic text data and corresponding sequences of typing data included in the set of input data;
• generating, using a character-level model, a character-level prediction for the set of input data, wherein the character-level model predicts an intended sequence of characters based on the text data and a sequence of typing actions included in the set of input data;
• determining, using a machine learning model, whether the set of input data detected at the computing device is authentic input, wherein the determining is performed based on the string-level prediction and the character-level prediction;
• and
• transmitting, to the computing device, an authentication decision, wherein the authentication decision is generated based on determining whether the set of input data is authentic.