Cybersecurity Statistics

VendorsAnvilogic

Anvilogic

Cybersecurity reports and statistics published by Anvilogic

0 categories1 reports

Research Reports

Reports and publications from Anvilogic

2025 State of Detection Engineering

Recent Statistics & Reports

36% said Streamlining workflows is a detection engineering program area that needs the most improvement.

88% believe AI will impact detection engineering in the next 3 years. 45% are using AI in their detection engineering efforts today.

45% said Alert enrichment is a skill that needs development for their detection engineering workforce.

45% said Reducing false positives is a detection engineering program area that needs the most improvement.

71% said Resource and time constraints is a challenge of building/maintaining custom detections.

42% of detections are custom-built, 37% are vendor-provided, and a smaller percentage are from open-source. Only 2% rely solely on vendor-provided detections.

Most detection engineers fall into the mid-career range.

67% said Behaviour-based detection type is the most effective.

67% of detection engineers reported strong leadership buy-in.

41% said Signature-based detection type is the most effective.

41% said Lack of skilled personnel is a challenge of building/maintaining custom detections.

76% said Understanding/mapping attack frameworks is the most valuable skill for their detection engineering workforce.

43% said Correlation-based detection type is the most effective.

54% said Lack of flexibility in customisation is a challenge of vendor-provided detections.

43% of detection engineers are using AI primarily for anomaly detection.

34% said Anomaly-based detection type is the most effective.

93% of organisations are using or planning to implement automation in their workflows.

43% said Improving turnaround time for developing and deploying a detection is a detection engineering program area that needs the most improvement.

34% said Collaboration between teams is a detection engineering program area that needs the most improvement.

43% said Threat intelligence-driven detection type is the most effective.

Showing 1-20 of 45 results

Top Categories