Google claims its machine learning models can now detect 99.9% of Gmail spam and phishing messages. These models integrate with the Google Safe Browsing tools for detecting links to malicious URLs. In addition, Google offers filters for use by enterprises to prevent unintended data sharing.