ARTICLES
Analysis, research, and perspectives on AI visibility, adoption, security, and model-agnostic capability management.
9 resources in articles
Companies are asking everyone to use AI while struggling to see what their teams have built. Visibility reveals what deserves more investment, what should become a shared standard, and where accountability is missing.
Read resourceWe tested AI agent skill security across 250 cases and compared deterministic scanning with an LLM-driven security review Skill, revealing where context, false positives, and risk composition matter most.
Read resourceOrganizations are moving from ad hoc AI conversations toward reusable capabilities that encode how work gets done. Those capabilities increasingly need ownership, versioning, governance, portability, and visibility.
Read resourceWhat the evidence says about AI skill performance, versioning and review, with broader workplace AI data clearly separated from skill-specific findings.
Read resourceAn AI skill registry connects reusable skills to their owners, sources, versions, and review history. Learn how it differs from a repository and when teams need one.
Read resourceSecurity and IT need more than a list of approved AI applications. Commonset provides an organizational layer for understanding which reusable AI capabilities exist, who owns them, which versions are trusted, who can use them, and where they are distributed.
Read resourceAI providers will continue to evolve, but the organizational capabilities built on top of them should not have to be recreated every time the underlying model, platform, or format changes.
Read resourceIf registering an AI capability makes it harder to use, employees have an incentive to keep it invisible. A better model separates visibility from approval and lets restrictions follow actual risk.
Read resourceAs reusable AI capabilities spread across models, providers, teams, and workflows, governance cannot live inside any one model platform. Organizations need a durable layer for ownership, versions, provenance, policy, access, approval, and trust.
Read resource