Sophos AI is working to make the web a safer place, using deep neural networks to detect malicious URLs, detect and warn users about phishing sites, and block malware delivery at the source. These deep learning models add additional layers of security to Sophos’s synchronized security architecture, stopping threats before they ever reach your network.
With the rise in prevalence of more complicated cyber-attacks, such as polymorphic malware, scripting and other living-off-the-land attacks, it has become easier to bypass traditional file scanning based anti-virus defenses. To protect against this evolution of malware, we are exploring new approaches to detection, such as behavior analysis, learning on graphs, and contextual models.
One of the best-known secrets of machine learning (ML) is that the most reliable way to get more accurate models is by getting more training data and more accurate labels. Learn about the ways that our engineering team build systems to handle the volume of data required for our state-of-the-art security model.
Our interpretable ML research focuses on inventing, prototyping, and operationalizing methods to explain the “thought processes” of our security machine learning systems. This work has resulted in multiple successful commercial model shipments and patents.
Social engineering attacks are on the rise, and remarkably resistant to detection by traditional signature-driven approaches. We’ve taken state-of-the-art work in natural language processing and adapted it to detect phishing and BEC email with remarkable detection rates.