Google unveiled its Secure AI Framework (SAIF) on Thursday to help the global artificial intelligence (AI) community collaborate, build, and develop the emerging technology under mutual standards.

Announced at the Aspen Security Forum, the Coalition for Secure AI (CoSAI) was born, following nearly a cumulative year of development.

Today at the Aspen Security Forum, alongside our industry peers, we’re introducing the Coalition for Secure AI (CoSAI). Learn more about the first three areas of focus the coalition will tackle ↓ https://t.co/orAdNod6aE

— Google (@Google) July 18, 2024

Under the new framework, coalition partners aim to focus on security measures that address current and forecasted AI risks.

The Coalition explained that its founding members — Amazon, Anthropic, Chainguard, Cisco, Cohere, GenLab, IBM, INtel, Microsoft, NVIDIA, OpenAI, Paypal, and Wiz — and many others will work together under the OASIS Open global standards and open source consortium.

Furthermore, the CoSAI Coalition will work with other prominent tech consortiums such as the Partnership on AI, Open Source Security Foundation, the Frontier Model Forum, and Machine Learning (ML) Commons amid its efforts to guide and direct “responsible” AI.

CoSAI’s Three Focus Areas

According to the initiative, it aims to take on three key focus areas:

1. Software Supply Chain Security for AI Systems, where Google’s efforts to “[extend] SLSA Provenance to AI models.” This allows others to identify whether the AI software is secure “by understanding how it was created and handled throughout the software supply chain.”
2. Preparing defenders for a changing cybersecurity landscape by providing a workstream to identifying “investments and mitigation techniques” to determine the effects of using certain AI.
3. AI security governance by developing “a taxonomy of risk and controls, a checklist, and a scorecard” to help boost readiness assessments, monitoring, and reporting AI product security.

Meta and IBM Open AI Alliance

The news comes after Meta Platforms and IBM launched their AI Alliance in December last year. In the hopes of securing an ethical framework for AI, Alliance has united more than fifty founding members and partners to build “open innovation and open science in AI,” it said at the time.

Companies and organisations partnered with the Alliance include AMD, Boston University, University of California Berkeley, CERN, the Cleveland Clinic, Cornell University, Dartmouth, Dell Technologies, Harvard University’s MOC Alliance, Hugging Face, University of Illinois, Imperial College London, Intel, INSAIT, the Linux Foundation, the University of Notre Dame, Oracle, Red Hat, Service Now, Sony, University of Tokyo, Yale University, and many more.