Closing the AI Trust Gap: Telecoms Launch New Certification Standard
London, Tuesday, 6 October 2026.
While 72% of telecom providers believe their artificial intelligence is trustworthy, only 14% can prove it. A new global certification initiative aims to bridge this critical operational gap.
Strategic Launch of Trustworthy AI Initiative
On October 6, 2026, industry alliance TM Forum announced a strategic collaboration with Accenture (NYSE: ACN) to launch a multi-year Trustworthy AI program [1]. This initiative aims to establish the industry’s first official AI trust certification, addressing operational risks and governance challenges as telecom executives seek to scale generative AI securely [1]. The program is designed to accelerate autonomous operations across the global telecommunications sector by providing a common blueprint for trust [1]. Guy Lupo, EVP of Trustworthy AI and Data at TM Forum, stated that trustworthiness only becomes operational when every principle is supported by continuous, verifiable evidence [1]. Arnab Chakraborty, Chief Responsible AI Officer at Accenture, noted that agentic AI marks a fundamental shift from AI supporting decisions to AI making decisions [1].
Quantifying the Industry Trust Gap
TM Forum research indicates a significant discrepancy between perceived and verifiable AI trustworthiness within the telecom industry [1]. While 72% of Communication Service Providers (CSPs) believe their AI capabilities are trustworthy, only 14% possess the ability to verify these claims [1]. This creates a trust gap of 58 percentage points between belief and verification capability [1][4]. TM Forum is a London-based nonprofit alliance founded in 1989, currently comprising over 800 member organizations including top-tier CSPs and hyperscalers [4]. The joint initiative aims to deliver a common benchmark for governance maturity and operational accountability to close this gap [4].
Operationalizing Trust Through Hackathons
To demonstrate practical application, a Trustworthy AI & Data Hackathon commenced on October 5, 2026, in Dallas, Texas [3]. The event involved over 200 hackers, with 110 participants on-site and over 100 participating virtually, totaling a minimum of 210 attendees [3]. More than 40 teams were tasked with building AI agents that provide a full chain of evidence for AI agent operations [3]. ServiceNow partnered with TM Forum to provide an end-to-end sandbox environment for the hackathon alongside NVIDIA, AWS for Telecom, Accenture, and Dell Technologies [2]. The technical infrastructure included AWS dedicated sandboxes featuring AI-native Open Digital Architecture (ODA) Canvas and a tamper-evident audit ledger [3].
Innovate Americas and Future Outlook
The launch coincides with the Innovate Americas event taking place in Dallas from October 6 to October 7, 2026 [4]. The event features over 29 exhibitors and various industry partners including Ericsson, Oracle, and IBM [4]. A networking reception sponsored by Tata Consultancy Services (TCS) was scheduled for October 6, 2026, from 17:30 to 18:30 [4]. TM Forum CEO Nik Willetts described AI as a gift to the industry that provides a license to reimagine the business from the top [4]. The competition aimed to identify the most secure and auditable solution, though the final winner status remains pending as of the announcement date [3][alert! ‘Winner status unknown at time of writing on 2026-10-06’].