Sójka will aid defend children online. How? Model makers explain

On CyberDefence24 we reported about [Sójk's premierz](https://cyberdefence24.pl/cybersecurity/new-from-white-soy-discovers-harmful-content). It is simply a tool that detects dangerous content from 5 categories: aggression, profanity, sexual content, criminal activity, self-harm. In real time, it analyses promotions and answers, and can besides support the moderation of content on social media. White defender codenamed Sójka is based on Polish artificial intelligence from [Bielik AI](https://cyberdefence24.pl/technologie/polski-bielik-on-great-scen-ai-wpolpracie-z-nvidia).