Google recently unveiled Magika, a groundbreaking open source tool powered by artificial intelligence that is designed to identify file types quickly and accurately. This innovative tool is now part of the AI Cyber Defense Initiative cybersecurity project, with the goal of providing information security specialists and other users with more effective automated tools.
One of the key challenges for services that work with files is accurately determining their contents. Traditional methods, such as analyzing file extensions or using heuristic rules, often fall short and are prone to errors. Magika, on the other hand, utilizes trained models to rapidly and precisely identify file types based on their contents. This technology has already been incorporated into various Google products like Gmail, Google Drive, safe viewing in Chrome, and Virustotal.
Compared to conventional systems that rely on manually configured rules, Magika technology boosts the accuracy of file type identification by 50% with a remarkable precision rate of 99%. However, despite its high accuracy, the system still makes classification errors in around 3% of cases.
As part of the AI Cyber Defense Initiative, Google is also planning to collaborate with 17 startups from the UK, the USA, and Europe to train automated tools and enhance cybersecurity measures. Google leaders stress that artificial intelligence is a crucial asset that can significantly improve cybersecurity defenses. They view Magika and similar technologies as key components in addressing global security threats and moving toward a secure and reliable digital environment.
Therefore, Google is not only bolstering its own products through the advancements in artificial intelligence, but also striving to elevate cybersecurity on a broader scale by offering its innovations for the greater good of the public.