Bash:

 find -type f -printf "%T@ %p \n" \
     | sort  \
     | tail -n 1  \
     | sed -r "s/^\S+\s//;s/\s*$//" \
     | xargs -iSTR cp STR newestfile

where "newestfile" will become the newestfile

alternatively, you could do newdir/STR or just newdir

Breakdown:

1. list all files in {time} {file} format.
2. sort them by time
3. get the last one
4. cut off the time, and whitespace from the start/end
5. copy resulting value

Bash:

 find -type f -printf "%T@ %p \n" \
     | sort  \
     | tail -n 1  \
     | sed -r "s/^\S+\s//;s/\s*$//" \
     | xargs -iSTR cp STR newestfile

where "newestfile" will become the newestfile

alternatively, you could do newdir/STR or just newdir

Breakdown:

1. list all files in {time} {file} format.
2. sort them by time
3. get the last one
4. cut off the time, and whitespace from the start/end
5. copy resulting value

Important

After running this once, the newest file will be whatever you just copied :p ( assuming they're both in the same search scope that is ). So you may have to adjust which filenumber you copy if you want this to work more than once.