In short
A data lake is a data architecture in which structured and unstructured data are stored in raw form in a scalable repository and processed when needed.
A data warehouse organises data into predefined schemas, whereas a data lake stores data first and applies the schema on read. This flexibility suits diverse sources such as sensor data, logs and documents.
Without governance a data lake can easily become a data swamp. A catalogue, access controls, data quality rules and lineage tracking are needed to keep it usable.