Improve bot and abuse detection accuracy

When you analyze the traffic reaching your origin after applying bots and abuse security controls, you may identify potential false positives (legitimate traffic misclassified as malicious) or false negatives (malicious traffic misclassified as legitimate). To help us improve detection accuracy, report your discovery in the Web Security Analytics Statistics view.

📘

Mind that:

• This is not an escalation mechanism for real-time attack mitigation,
• During an attack, you should contact Akamai Support,
• Your feedback is processed and evaluated, but no follow-up notification or status update is issued.

1. Go to Web Security Analytics.

2. Set the following filters:
   One of these operation filters:

   • Common > API Operation,
   • Common - HTTP Info > Hostname and Path.

   At least two of the following filters:

   • Common - Client info > Client TLS Fingerprint V2,

   • Common - Client into > Client TLS Fingerprint V3,

   • Common - Client info > JA4 Client TLS Fingerprint,

   • Common - Network info > Connecting IP Address,

   • Common - Network info > End User IP Address,

   • Common - Network info > Connecting AS Number,

   • Common - Network info > End User AS Number.

   • Account Protection > User ID,

   • Account Protection > Device ID.

   📘

   You can see the Detection issues link only when you select these filters, and only if the period you're looking at is 7 days or less. Otherwise, it doesn't appear.

3. In the Statistics view, if you notice any inaccuracies, mouse over the filter settings, then click Detection issues?
   The following image shows the Detection issues? link in the Statistics view:

4. In the window, select the following details:

• Type:
  False positive or false negative.
• Subtype (for false negatives only):
  Loyalty program abuse, payment abuse, carding, inventory hoarding, promotion abuse, SMS abuse, scraping, login abuse, account opening abuse, bot abuse.

5. Click Submit.