TM Forum takes AI trust from promise to proof at world-first industry Hackathon

More than 200 developers across 40 teams competed in TM Forum’s Trustworthy AI & Data Hackathon, advancing practical methods for trusted AI.

LONDON, UNITED KINGDOM, October 14, 2026 /EINPresswire.com/ — TM Forum today named the winner of a world-first industry Hackathon that brought together more than 200 developers across 40 teams and demonstrated practical pathways for turning trust in autonomous AI from an ambition into engineered proof in a real network environment.

TM Forum’s Trustworthy AI & Data Hackathon: the quest for proof of trust – took place during Innovate Americas (October 6-7 in Dallas), challenging developers to show that agentic decisions can be controlled, explained and verified without slowing deployment.

The Hackathon winner – Infyvibes, a team from Infosys – turned governance requirements into verifiable operational assurance, and advanced further than other competing teams to demonstrate how autonomous AI operations can be stopped and routed for human approval when defined limits are breached.

For three days, developers at the TM Forum event in Dallas, and joining virtually from around the world, worked inside a live autonomous enterprise sandbox. They configured the TM Forum AI-Native Open Digital Architecture (ODA) Canvas, connected agents and tools, and tested governance controls under sustained – sometimes hostile – activity. The result was a world-first advancement in engineering trust into autonomous operations, and large-scale, hands-on validation of how TM Forum assets can support real-world trusted autonomy.

“Autonomy without proven assurance is a leap of faith the industry cannot afford,” said Guy Lupo, EVP, Trustworthy AI & Data, TM Forum. “We need trustworthiness built into the core code, systems and processes that AI depends on, so trust becomes an engineering discipline rather than a bolt-on compliance exercise. Our ground-breaking Hackathon put this principle to the test at scale, in the hands of developers, on a real telecommunications challenge.”

The response from the industry showed that this is more than a lab exercise. Participants created 196 additional agents and executed more than 35,000 governed model calls during the challenge. Across 50 running ODA Canvas instances, teams configured components, integrated agents and used Canvas documentation and repositories to build working solutions. The activity provided substantial hands-on validation of the AI-Native ODA Canvas and trained a new community of developers to build with ODA.

TM Forum’s Model-as-a-Service (MODaaS) acted as the governing gateway for model traffic throughout the Hackathon, applying policy, identity, authentication, encryption and guardrails in the AWS environment to demonstrate outstanding performance under real-life operational load.

The winning team that proved trust at runtime

The Trustworthy AI Hackathon was supported by Accenture, AWS, Dell Technologies, NVIDIA and ServiceNow, hosted by AT&T, and deployed on TM Forum’s ODA Canvas. Together, this created a sandboxed environment in which governance, risk and compliance could be securely tested against live evidence of agent behavior across cloud and enterprise operations.

Infyvibes demonstrated all three of the challenge’s required controls for spend, event recording and performance drift, turning governance requirements into verifiable operational assurance. The team went further to implement real-time enforcement and a human-in-the-loop escalation to show how operators can increase autonomy while retaining accountability and control.

Other teams extended the challenge in unexpected ways, including building an intrusion-detection capability for agents from infrastructure audit information and connecting it to a compliance dashboard. These contributions showed how quickly developers can move from learning the environment to strengthening it.

What happens next?

TM Forum’s Trustworthy AI and Data Mission champions trust as an enabler of autonomy at scale. But autonomy cannot scale if every operator builds a different trust model. Global customers, regulators and partner ecosystems need a common industry approach to trustworthy AI.

This Hackathon is the Mission’s first practical step towards common methods, standards and industry assets for governing AI-native operations. TM Forum will now take the code, evidence, controls and lessons from Dallas to its global membership, turning working practice into a shared industry approach to trusted AI and autonomy.

TM Forum’s next hackathon takes place at Accelerate 2027, February 1-4, in Amsterdam.

About TM Forum
TM Forum is an alliance of over 800 organizations spanning the global connectivity ecosystem, including the world’s top ten Communication Service Providers (CSPs), top three hyperscalers and Network Equipment Providers (NEPs), vendors, consultancies and system integrators, large and small.

We provide a place for our Members to collaborate, innovate, and deliver lasting change. Together, we are building a sustainable future for the industry in connectivity and beyond.

To find out more, visit: www.tmforum.org

Media Contacts
newsroom@tmforum.org

Katy Branson
TM Forum
newsroom@tmforum.org

Media gallery