United States

TM Forum and Accenture announce AI certification plan for telecom operators

The programme aims to test autonomous-system decisions against auditable evidence. Its certification is still to be developed and does not establish current deployments.

TM Forum announced a multi-year programme with Accenture on October 6 at Innovate Americas in Texas to define how telecom operators can assess trust in their artificial intelligence systems. Its planned outcome is an AI governance certification examining controls, workforce readiness and operational accountability. The announcement starts joint work; it does not mean that any company has already received the certification or that a new network is now running autonomously. [1][2]

The industry problem is concrete. A system may detect a network anomaly and recommend a repair, but letting it carry out changes requires a record of the data it used, the decision it made and who can stop or reverse it. TM Forum proposes verifiable evidence of interactions between AI agents and their effects. Telecoms Tech News notes that such controls also matter in customer service, billing and service assurance, where mistakes can affect subscribers and contracts. [1][2]

The initiative aims to provide common criteria for comparing AI governance across communications service providers. TM Forum says the assessment will cover the maturity of controls, staff preparation and accountability. Accenture will help turn those criteria into operational practices. Final requirements, audit procedures, certification timing and adoption by individual operators have yet to be publicly specified. The announced programme must therefore be distinguished from a certification already on offer or measured technical results in live networks. [1][2]

TelecomTV included the announcement in its weekly review of industry AI news and highlighted the search for evidence about autonomous decisions. Independent coverage confirms the stated scope, but still relies heavily on TM Forum's account: it does not establish that the future controls work or that operators have adopted them. The sector test will be whether the rules allow specific actions to be proved and reviewed before AI is allowed to change networks or services. [1][2][3]

An earlier survey cited by TM Forum found a gap between operators' confidence in their AI and the evidence they could provide for it. That is background to the announcement, not a new measurement made on October 6. The next verifiable advance will be published criteria and audits that demonstrate a real certification. The accompanying picture of a network operations centre is an archive image from 2008; it does not show TM Forum's programme or a current AI system. [1][2]

Report an error →