The short explanation of "copyleft" licenses is that if you apply a "copyleft" license to your code, then anyone who creates a derivative work from your code must apply the same license to their code. This implies that any derivatives of the derivative, etc also will have the same license. In essence, it keeps your code and anything derived from it open source. Whereas with an all permissive license like MIT, there is no such obligation on those who create derivatives (e.g., MIT licensed code can be used by others, including derivatives, in closed source software). In both cases, you are still the copyright owner, but you license away more of your rights with MIT or Apache than you do with the GPL and other "copyleft" licenses.
The short explanation of "copyleft" licenses is that if you apply a "copyleft" license to your code, then anyone who creates a derivative work from your code must apply the same license to their code. This implies that any derivatives of the derivative, etc also will have the same license. In essence, it keeps your code and anything derived from it open source. Whereas with an all permissive license like MIT, there is no such obligation on those who create derivatives (e.g., MIT licensed code can be used by others, including derivatives, in closed source software). In both cases, you are still the copyright owner, but you license away more of your rights with MIT or Apache than you do with the GPL and other "copyleft" licenses.
Here's a link that explains in more detail: gnu.org/licenses/copyleft.en.html