Guozhen AIGlobal AI field notes and model intelligence

Realtime AI News

Vodafone and Microsoft call for common standards on AI agent skills

Vodafone and Microsoft have publicly called for standards governing AI agent skills, TM Forum's Inform reported on November 12. The report states the position but not the content or destination of any proposal, as operators move agents into network operations and customer service.

Published
沃达丰与微软呼吁为AI智能体技能建立统一标准
Image source: myvi.in

TM Forum's news outlet Inform published a report on November 12 under a headline that states the news plainly: Vodafone and Microsoft are calling for standards covering AI agent skills.

What the report makes public so far is the position rather than the specification. The two companies publicly back a common standard for AI agent skills, but the available summary does not list what such a standard would contain, nor where the proposal would land — whether it goes to a standards body for formal work, or starts as a shared way of describing skills inside their own platforms.

Skills are a natural place to start because of how agents work. To finish a task, an agent has to discover which tools are available, understand the inputs and outputs each tool accepts, and decide the order of calls. If every vendor defines skill names, parameters and capability boundaries differently, reusing an agent across platforms means rebuilding the integration each time.

That is especially concrete in telecoms. Operators are putting agents into network operations, customer service and BSS/OSS processes, with equipment vendors and cloud providers both involved. TM Forum has long pushed standardisation work such as open APIs and autonomous network levels, and agentic operations already features in its proof-of-concept research.

The pairing is worth noting too: Vodafone is a major operator in Europe and Africa, while Microsoft both supplies cloud infrastructure and sells Copilot-style agent products. Operators worry about lock-in to a single platform, platform vendors would like their own description of skills to become the de facto one, and both are now asking for a "standard" — not necessarily the same thing.

What to watch is whether the call enters a formal process: whether TM Forum sets up a working group or issues a specification, whether the work connects to other agent interoperability efforts, and whether operators start writing skill standards into procurement requirements.

Why it matters

Without shared definitions of agent skills, reuse across platforms means repeated integration work, and operators become easier to lock into a single vendor during procurement.

VodafoneMicrosoftAI AgentStandards
Back to realtime news

Nearby Updates

All

11/12, 15:45

Baidu World 2026: Baidu Unveils AI Results as Robin Li Says the Next Act Is the 'Intelligent Organization'

At its Baidu World 2026 conference, Baidu released a batch of AI results and founder Robin Li framed the next act of AI as the "intelligent organization." The messaging shifts attention from raw model capability toward how organizations themselves are restructured, though no specific product list or metrics were disclosed.

11/12, 13:34

ByteDance reportedly preps a Muse-style AI agent, with AI spending aimed at $70B

Dealroom reports that ByteDance is preparing a Muse-style AI agent while targeting roughly $70 billion in AI spending, without naming the product or giving a timeline. The report places the Chinese company in the general-assistant race against OpenAI and Meta.

11/12, 13:28

Chainalysis flags an on-chain shadow market behind China's model-distillation wave

Chainalysis has published an analysis describing an on-chain shadow market tied to China's AI model-distillation wave, with crypto settlement used to work around US export controls. The report connects on-chain payments to the transfer of model capability, pointing at the part of the supply chain controls see least clearly.

11/12, 10:33

Naver in talks for Asia's largest Nvidia server order, signs 2 trillion won AI deal

Naver said on its third-quarter earnings call that it is negotiating what would be Asia's largest server purchase order with Nvidia to secure Vera Rubin GPUs in advance. The company has also signed a roughly 2 trillion won computing services contract with a U.S. AI inference firm and expects its AI factory business to post operating profit from 2027.