Multiple imputation would be an option: https://www.ibm.com/support/knowledgecenter/en/SSLVMB_24.0.0/spss/mva/idh_idd_mi_variables.html.

I agree with dr Violeta

SPSS has the option of multiple imputation to deal with missing data.

Look it up at SPSS> analyze>multiple imputation> analyze patterns or > impute missing values.

you can use real statistics tool to deal with missing data

check it up at http://www.real-statistics.com/handling-missing-data/multiple-imputation-mi/.

The question was about using AI to make the imputation. to my knowledge that does not exist in SPSS, but I can coded.

http://www.real-statistics.com/handling-missing-data/multiple-imputation-mi/.

I think that Stata (www.stata.com) is becoming one of the leading packages; it is certainly the major rival to SAS and SPSS, and it is superior in at least some ways. It has a very interesting combination of a commercially developed core with the option of adding user-developed routines. If I were picking a new package now and were willing to pay the price, I'd definitely pick Stata over SAS or SPSS. I use SYSTAT (www.systat.com), which is an excellent value compared to most packages, but does not have the range of features you find in Stata or SAS. A problem with SPSS and some other packages is that they want to sell it to you piece by piece, so if you want a certain feature you may have to add it on as an expensive module. Most packages have limited trial versions available on their websites you can try. SYSTAT also has a free version for students called MYSTAT based on an older release with only minor limitations (limited to 100 variables, and lacking some of the more advanced statistics). In addition to R, there are other free packages out there, such as PSPP (http://www.gnu.org/software/pspp/), which is roughly speaking a free version of SPSS, and MicrOsiris (www.microsiris.com/).