This is hard to say. I lack any first-hand experiences, so I need to rely on vague cues. There are some good signs for them not being predatory, though, such as the following:
Their Twitter account shows that they have a relatively large following with a high degree of community engagement, being tagged by influential researchers and institutions such as the Physics Magazine (of the American Physical Society) etc. This indicates that they are taken seriously within a larger scholarly community. (However, their main account @theSAIorg seems to be inactive.)
Most of their editorial board members seem to know how to publish well.
They deposit metadata at CrossRef, according to a search for the publisher "Science and Information Organization" in CrossRef's Title List. (But then again, according to the discussion here, there are notorious predatory publishers that do the same.)
Their articles seem well-cited. Google Metrics shows a h5-median of 59 for this journal, for example.
Researchers affiliated at some reputational institutions publish in their journals. Just look at the left-hand side list titled "Top Institutions" at Microsoft Academic, such as for this journal (IJACSA) (e.g. showing the Max Planck Society) or for this journal (IJARAI) (e.g. showing the MIT).
At the same time, there are some puzzling circumstances, which makes me skeptical of my initial judgement. Such as the following:
- They are not indexed at the DOAJ, even though their journals are all Open Access.
- There are "< 10" peer review reports validated by Publons for their main journal, which could mean anything from 1 review to 9. And that is quite a low number for a journal that has published so many articles over multiple years.
- Likewise, there are no peer review reports documented at SciRev.
Thus, my conclusion is: I am not sure...