Interoperate
Can I connect?
Open interfaces connect agents, tools and services across models, vendors and ecosystems.
Explore this level ↗Open systems. Shared understanding. Human control.
Standards for AI.
Agency for people.
Building an open, understandable, trustworthy and human-governed AI ecosystem.
IEEE Computer Society / Artificial Intelligence Standards Committee
A standards architecture01 — 04
01InteroperateCan I connect?02UnderstandCan we understand each other?03TrustCan I rely on you?04GovernWho is ultimately in control?A proposed portfolio framework · September 2026
The AI divide is changing
Beyond access.
Toward agency and control.
Powerful models are only the beginning. The next questions are whether we can connect, share meaning, verify behavior and retain authority over the intelligence we use.
Understand
Can we understand each other?
Shared semantics let systems exchange meaning, context and intent—not just messages.
Explore this level ↗Trust
Can I rely on you?
Evaluation, evidence and clear responsibility make AI behavior verifiable and accountable.
Explore this level ↗Govern
Who is ultimately in control?
People and institutions retain authority over knowledge, identity, policies, skills, memory and decisions.
Explore this level ↗Proposed framework for committee discussion; project mappings are illustrative.
Portfolio update · 09 September 2026
From direction to progress.
New approvals
12
Newly approved PARs
The liaison report records twelve approvals, with eight projects highlighted for SC 42 coordination.
Explore highlighted projects ↗Emerging work
07
Proposals under review
New work on AI observability, GenAI lifecycle governance, world-model evaluation and more.
See the review snapshot ↗International coordination
SC 42
A shared standards conversation
Coordination topics connect interoperability, semantics, evaluation, trust and governance.
Explore coordination themes ↗Featured initiative · Agentic AI
Make the ecosystem
work together.
Agents need more than a connection. They need identity, shared context, delegated authority and evidence that travels with their work.
Discover & connectDescriptions · Registries · Interfaces
Understand & actMeaning · Context · Delegation
Verify & governProvenance · Evidence · Authority
An illustrative view of coordination themes.
Strategic initiatives
One ecosystem.
Connected areas of work.
Semantic interoperability
Align concepts, tasks, evidence and outcomes using terminology, ontologies, semantic profiles and schema mappings.
Explore initiative ↗Trust, observability & evidence
Connect model invocations, delegated actions, tool calls and outcomes to traceable evidence and operational assurance.
Explore initiative ↗Edge, distributed & physical AI
Connect edge-to-cloud execution, cross-platform deployment and manufacturing systems across trust domains.
Explore initiative ↗Standards as products
From specification
to implementation.
A practical standard needs an ecosystem: profiles, reference implementations, SDKs, conformance tests and feedback from real use.
Explore the implementation approachSpecify & build
Clear profiles, reference code, adapters and examples.
Test & connect
Conformance evidence and independent interoperability.
Adopt & improve
Plugfests, implementation feedback and measurable value.
The long-term question
Who controls your
accumulated intelligence?
Cognitive sovereignty means retaining control over knowledge, identity, policies, skills, memory and decisions as models and providers change.
Research / explorationEvents & community
Build the next chapter together.
Nov032026
Upcoming summit
Visit summit websiteAISC AI Standards Summit
Seoul, Republic of Korea · 09:00–13:00 KST
Daisy Room, 5th floor