Liaison statement
LS on SG17 new work item X.LLMCC: “Guidelines for Large Language Model data security based on Confidential Computing”

Additional information about IETF liaison relationships is available on the IETF webpage and the Internet Architecture Board liaison webpage.
State Posted
Submitted Date 2026-02-02
From Group ITU-T-SG-17
From Contact tsbsg17@itu.int
To Group SEC
To Contacts Paul Wouters <paul.wouters@aiven.io>
Deb Cooley <debcooley1@gmail.com>
Cc The IETF Chair <chair@ietf.org>
Scott Mansfield <Scott.Mansfield@Ericsson.com>
Deb Cooley <debcooley1@gmail.com>
Paul Wouters <paul.wouters@aiven.io>
Response Contact arnaud.taddei@broadcom.com
jhnah@etri.re.kr
gaofeng149@chinaunicom.cn
Technical Contact arnaud.taddei@broadcom.com
Purpose For information
Attachments Output – Proposal for new work item X.LLMCC: “Guidelines for Large Language Model data security based on Confidential Computing”
Body 
ITU-T SG17 is pleased to inform you that SG17 established a new work item:
draft new Recommendation ITU-T X.LLMCC: “Guidelines for Large Language Model
data security based on Confidential Computing” (SG17-TD165-R1/WP4 - Attached)
at the ITU-T Study Group 17 plenary meeting (Geneva, 3-11 December 2025).

Scope

This Recommendation provides data security guidelines for Large Language Model
enabled by confidential computing technologies. It specifies principle-based
controls to protect the confidentiality and integrity of data across the Large
Language Model lifecycle.

This Recommendation is applicable to AI system developers, Large Language Model
providers, cloud service providers, and third-party solution providers to
implement Confidential Computing-enabled security enhancement solutions.

Summary

The widespread deployment of Large Language Model services, particularly in
cloud environments, faces escalating security threats such as adversarial
attacks, data leakage during computation, and model tampering. Traditional
security measures inadequately address risks to "data in use," while fragmented
implementations of Confidential Computing technologies lack cross-platform
interoperability and trust in multi-party collaborations.

This Recommendation establishes harmonized guidelines to secure Large Language
Model through Confidential Computing, focusing on system-layer isolation,
encrypted model execution, and algorithm integrity. It addresses critical gaps
by unifying implementations across vendors, defining attestation protocols for
verifiable trust, and providing layer-specific safeguards for hardware, data,
and models.

Applicable to AI system developers, model providers, and cloud service
providers, the proposal standardizes performance-optimized protections for LLM
workflows without overlapping with general AI security standards to enable
secure LLM adoption at scale.

ITU-T SG17 will keep you informed of its progress and looks forward to
continued collaboration with ITU-T SG21, ISO/IEC JTC 1/SC 27/WG4 and IETF on
Confidential Computing for LLM related topics.