The Linux Foundation has formed the Appia Foundation to provide modular specifications and to bridge the global standards with trusted assessments in the AI value chain.

Hosted under the Joint Development Foundation (JDF), Appia Foundation was created to address the challenges that come with changing regulations and standards.

With regulations moving to active enforcement, more value chain partners require evidence of trustworthy AI in contracts and vendor evaluations.

Meanwhile, international frameworks like ISO/IEC standards establish vital foundations, with the potential to unlock opportunities for a consistent, shared mechanism to translate these global rules.

Against this backdrop, Appia Foundation aims to provide an open connecting layer by building on international standards and established frameworks to develop publicly available global specifications.

The Linux Foundation said that the Appia specifications are organized across a Requirements and Guidance layer and an Assessment Enablement layer to offer the testing criteria, evaluation guidelines, and component typologies needed to effectively assess AI applications and processes.

“AI systems now make decisions about people's loans, their children's schools, and their jobs. People on the receiving end deserve to know those systems were built and assessed against criteria that hold up to scrutiny,” Craig Shank, executive director of the Appia Foundation, said.

“The Appia Foundation was formed to do that work: creating publicly available specifications that organizations across the AI value chain use to demonstrate their systems meet those criteria. By establishing this open framework, we are building the accountability layer required to scale safe and trusted AI across major industries.”

Some of the first members of the Appia Foundation include the likes of Ericsson, Google, Mastercard, Microsoft, Arm, Mitsubishi Electric, OpenAI, Schneider Electric, and Siemens.

Commenting on the development, Natasha Crampton, chief responsible AI officer at Microsoft, acknowledged that as AI adoption accelerates, “organizations need practical ways to demonstrate that their systems are trustworthy”.

“The Appia Foundation can help by translating emerging standards and governance expectations into more consistent, assessable evidence across organizations and jurisdictions. That kind of shared, practical foundation is important to building trust and helping AI scale more responsibly,” she added.

The Appia Foundation said that instead of assessing an entire system from scratch, organizations can evaluate only what is relevant to their role, component, and regulatory context.

When an upstream provider demonstrates conformity with relevant modules, that technical evidence passes through to downstream users. This enables seamless reuse of conformity evidence across the value chain, reducing friction while maintaining clear boundaries of accountability, Appia said.

Per Beming, chief standardization officer of Ericsson, said: “At Ericsson, we believe harmonized standards and regulations are essential to scaling trustworthy AI globally. The Appia Foundation helps establish conformity specifications that give industry a practical and consistent way to demonstrate conformity across the AI value chain for mobile infrastructure.”

A flurry of recent Linux Foundation projects

This month, the Linux Foundation also launched the OpenSharing project aimed at standardizing AI data and asset sharing alongside data analytics giant, Databricks.

The project looks to create vendor-neutral protocols for organizations sharing AI assets and data with Databricks, contributing to this effort.

The partners aim to create an open means to exchanging agent skills, AI models, and unstructured data volumes, leaning on learnings from the Delta Sharing protocol.

Earlier in June, the Linux Foundation also unveiled plans to launch the Tokenomics Foundation to standardize AI infrastructure spending with Google, Microsoft, Salesforce, and Oracle backing efforts to tame exploding AI token costs.

Several technology leaders have expressed initial support, with the newly created foundation set to work closely with the FinOps Foundation to bring the practice of managing variable tech costs to token-driven AI.