The Linux Foundation has taken custody of DNS-AID, an open source project for AI agent discovery that was originally built by network infrastructure vendor Infoblox. The move matters because it shifts a discovery protocol out of a single company's hands and into a governance structure with actual competitors at the table. The project's member list is the tell: Cloudflare, GoDaddy, Equinix, Internet Systems Consortium, and CSC — companies that would normally compete to own the naming layer for agentic infrastructure — are now co-stewards of the same spec, alongside Indeed, IDC, and WWT. Technically, DNS-AID doesn't invent new DNS record types or protocol values. It reuses existing SVCB and TXT records, layered with optional DNSSEC and DANE TLSA validation, to let organizations publish agent metadata under a well-known label in their own DNS namespace. The reference implementation ships as a Python SDK, a CLI, and an MCP server, so a developer can query for an agent by name, function, or domain and get back a signed answer describing which protocol to speak — MCP, A2A, plain HTTPS — without a central directory brokering the lookup. That design choice is the whole argument. Two competing project maintainers framed it as making agent discovery 'web-native' — reusing the reliability of DNS rather than building a new registry that someone has to run, trust, and eventually monetize. It's the same argument the IETF draft covering this exact mechanism makes in more formal language: no new DNS message structure, no new opcodes, just a namespace convention for capability advertisement. This sits next to, and largely validates, the academic case made in a recent arXiv paper on agent discovery, which frames the crowded field of competing discovery proposals as needing a shared evaluation framework rather than another one-off standard. DNS-AID is effectively an attempt to settle that question in practice before the academic debate settles it in theory: instead of a bespoke Agent Name Service or a proprietary registry, put the metadata in DNS, where resolution, caching, and security extensions already exist and are already operated at Internet scale by the same companies now backing the project. Whether it actually displaces the half-dozen other discovery drafts circulating in IETF and industry venues is unresolved. But getting Cloudflare, GoDaddy, and ISC into the same governance room is a stronger signal than another whitepaper would have been.