The reason might be that forward slashes are allowed to be escaped in JSON.
The forward slashes in the following JSON might actually be escaped with a backslash.
Unescaped:
{"foo/bar":"bar/baz"}
Escaped:
{"foo\/bar":"bar\/baz"}
When retrieving such escaped data back, the decoder needs to handle the \
in front of the /
and ignore them. I think most decoders do not care whether or not forward slashes are escaped, at least they parse JSON data happily with forward slashes being escaped and without.
In any case, when you retrieve your data back as a python object from the decoder, the escape characters should go away. Handling escape sequences is a task encoders/decoders should handle transparently for you.
To go on: can you veryify if the JSON response from the server is actually correct? If yes, can you then try with a standalone python program whether your decode handles encoding/decoding of such string correctly?