ETSI, Japan's TTC and South Korea's TTA sign Letter of Intent on 3GPP-style AI project
ETSI, Japan's TTC and South Korea's TTA signed a Letter of Intent in Seoul on Monday to explore a 3GPP-style AI Partnership Project, Tech Times reported on 3 November. The move comes as Brussels delays the AI Act's toughest obligations.

The Letter of Intent was signed at the Global ICT Standards Conference in Seoul. The theme was "AI for All". It commits ETSI, Japan's TTC and South Korea's TTA to explore creating the AI Partnership Project, or AIPP, and nothing more.
For anyone selling AI in Europe, the practical route to compliance is shifting from statute to sandboxes and industry specifications. That may cut short-term costs. It also leaves the legal duties unclear: neither the timing of the delayed high-risk rules nor the enforceability of any AIPP output is settled.
A 3GPP-style AI body takes shape
The template is 3GPP, the Third Generation Partnership Project. Seven regional telecoms standards bodies set it up in Copenhagen in December 1998. Its specifications now underpin networks used by more than five billion mobile subscribers. AIPP would copy the same organisational architecture and governance logic, in the hope of producing globally coherent technical specifications faster than intergovernmental bodies manage. ETSI Board Chair Markus Mueck, an Intel engineering director, framed the project as a complement to the existing AI ecosystem rather than a replacement for it.
"Artificial Intelligence is evolving at unprecedented speed, making global collaboration on standards more critical than ever," Mueck said.
Mueck said an AI Partnership Project could align priorities and accelerate standards development. The output would be technical specifications, not legislation.
The AI Act timetable slips
On the European side, the calendar has moved the other way. The EU recently delayed application of its high-risk AI rules and eased some administrative burdens, economy.ac reported on 28 October. No new date has been announced.
Investment has responded to the softer tone. Nvidia plans to commit USD 10 billion to companies, universities and research institutions across Europe. Of that, USD 250 million in computing resources is earmarked for AI safety research over five years. The company named four German startups as recipients at the Invest in Germany Summit in Berlin but did not disclose individual amounts. The scale gap remains the backdrop. ECB President Christine Lagarde said the US accounts for 75% of global AI computing capacity through its data centres, while Europe accounts for 5%. She also said the US developed 59 notable AI models last year and China 35, while the UK and France produced one each.
Sandboxes become the testing ground
With the statute in flux, sandboxes have become the place where the rules get argued over. Dentons published five proposals on 30 October for effective implementation of AI regulatory sandboxes. FinanceX Magazine examined what the AI Omnibus Regulation means for those same sandboxes the same day. Neither is an adopted rule.
The fragmentation they respond to is real. A company selling AI across borders already faces the AI Act and separate state laws in Colorado, Texas, California and New York. Add China's algorithm governance rules and the national frameworks emerging in Japan and South Korea. The 3GPP model offers less certainty than it appears to. 3GPP does not produce standards. It produces specifications, agreed by company members through consensus. Those gain legal force only when regional bodies such as ETSI, TTA and TTC transpose them into their own national or regional standards.
Compliance stays fragmented
AIPP has not published a single specification. The three signatories also lack answers on how the body would deliver globally coherent specifications, and membership, funding and voting weight are unresolved, according to Tech Times. The Letter of Intent commits the three bodies to explore creating AIPP, but it does not create it.
What happens next
Dentons' five proposals and FinanceX's analysis are inputs to the sandbox debate, not rules anyone has to follow yet. The EU has not said when the delayed high-risk provisions will apply.
For now, voluntary specifications and pilot programmes are doing the work. Enforceable rules were meant to do it.
Sources
18- 01Europe, Japan, and South Korea Propose Global AI Standards Body Modeled on 5G, Without US or China - Tech TimesEN
- 02“Regulatory Easing Energizes Investment”: Europe’s AI Ecosystem Gains Growth Momentum as Defense Startups Advance Amid RearmamentEN
- 03AI regulatory sandboxes: Five proposals for effective implementation - DentonsEN
- 04The AI Omnibus Regulation: What’s Next for the AI Regulatory Sandboxes? - FinanceX MagazineEN
- 05The TechCrunch Founder Summit agenda is only half the reason to be in Boston on November 4EN
- 06Tomorrow: TechCrunch Founder Summit takes over Boston and when pass rates increaseEN
- 07ARIA’s Scaling Inference Lab Deploys Lumai’s Optical Compute TechEN
- 08London-based LO:TECH lands $12M Series A led by MMCEN
- 09Palantir AI tech unused or little used by 29 NHS trusts said to be benefiting from itEN
- 10European tech weekly recap: Over €1.3B invested across 60+ dealsEN
- 11The biggest conversations in health techEN
- 12Inside Automotive DMS & OMS: Technology, Challenges & Power Solutions EN
- 13Forget AI, Europe has a plan to win the next tech revolutionEN
- 14How Beta Technologies Is Clearing the Runway for Electric FlightEN
- 15Canada to join European supercomputer program, collaborate on tech priorities - BetaKitEN
- 16AI practice note and what agentic AI means for legal: 7 technology stories you might have missed - The Law SocietyEN
- 17Media rights squeeze, US investment and AI tech: The sports business trends from Sportel 2026EN
- 18TechnoMile Recognized as a 2026 Inc. Power Partner Award WinnerEN
All figures and quotations in this text come from the sources listed below.
Content prepared by the editorial team with AI assistance.
Comments
0- No comments yet — be the first.