IAS's solution works both to prevent ads from appearing on sites with unsafe content and to block ads that have been delivered to those sites.
Its algorithm crawls the web's URLs in real time and classifies content into eight categories: adult content; alcohol; fake news; gambling; hate speech; illegal downloads; illegal drugs; offensive language; and violence. These categories are then split into four risk thresholds: low; moderate; high; and very high.
Phill Hayman, head of technical services, sales engineering at IAS explained: "These scores are offered to brands so they can decide which threshold they're comfortable with. A gambling client, for example, would have different ideas of brand safety than say, Pampers."
Until now however, IAS's solution has only included around nine European languages.
"Speaking to brands, they now need brand safety support globally, in pretty much every language on the planet," Hayman said.
The language expansion, developed in-house by IAS's team of data scientists and linguists now includes languages from Eastern Europe, the Nordics and Asia. In Asia, these now include Mandarin (both simplified and traditional), Japanese, Thai, Korean and Tagalog.
IAS has ambitious plans for China and this update should help it to continue to "grow aggressively", said Hayman.