|
Work item:
|
F.ACAIA-B
|
|
Subject/title:
|
Assessment criteria for artificial intelligence agents: Benchmark
|
|
Status:
|
Under study
|
|
Approval process:
|
AAP
|
|
Type of work item:
|
Recommendation
|
|
Version:
|
New
|
|
Equivalent number:
|
-
|
|
Timing:
|
2028-01 (Medium priority)
|
|
Liaison:
|
ITU-T SG20
|
|
Supporting members:
|
Beijing Baidu Netcom, China Mobile Communications Corporation.
|
|
Summary:
|
This Recommendation specifies a comprehensive framework for benchmarking AI agents. It provides a structured methodology for assessing and evaluating the general capabilities and task performance of AI agents in a consistent, reproducible, and comparable manner.
This Recommendation specifies a comprehensive framework for benchmarking AI agents. It provides a structured methodology for assessing and evaluating the general capabilities and task performance of AI agents in a consistent, reproducible, and comparable manner.
|
|
Comment:
|
-
|
|
Reference(s):
|
|
|
Historic references:
|
|
Contact(s):
|
|
| ITU-T A.5 justification(s): |
|
|
|
|
First registration in the WP:
2026-05-12 11:02:49
|
|
Last update:
2026-05-12 11:33:55
|
