I can define both but I keep second-guessing which one a scenario is describing.
Accepted answer
Lossless compression lets you reconstruct the original data **exactly** — nothing is discarded, it just stores the data more efficiently (PNG, ZIP). Lossy compression permanently throws away data judged less noticeable to get a much smaller file (JPEG, MP3), and you can never recover the original.
The scenario cue is usually the requirement, not the format: if the data must be perfectly restored (source code, a text document, a bank record), you need lossless. If the priority is a small file and some quality loss is acceptable (streaming audio, photos on a web page), lossy is the fit.