AWS WAF can be improved by not being too sensitive for some data because with some XML, it is sometimes very sensitive. When we upload videos or photos, it is flagged as a malicious pattern, and sometimes it is not sensitive enough because some query patterns that have the capability to perform malicious actions still result in false positives from AWS. It would be better for AWS WAF if it not only had logs but also specific payloads or something that directly included the specific logs so we can have more detail in our analysis. I think it should be better if they improve their capabilities in reducing the attack surface and also have better visibility for OWASP top 10 web applications and the API. Regarding AWS WAF's AI capabilities, its governance and security are very good because our attack surface is mostly reduced. The new dashboard is a bit complicated; it is easier for non-technical users, but for technical professionals, it is sometimes confusing. While the security itself is good and the governance is good, I think it needs to improve the UI or the flow itself. The accuracy and reliability of AWS WAF's output is better than others. With the free version, it has a good filter with the attack surface and the reliability is good, but sometimes we need to adjust. With the free version, we also have to pay some amount of money if the number exceeds the default provided by AWS.
