A new working group founded by IBM, Nvidia, and Red Hat under the Linux Foundation's LF AI & Data project has launched the DocLang Specification — an open, AI-native document format designed to make business documents machine-readable for LLMs. Unlike existing formats such as PDF or JPEG built for human consumption, DocLang aims to provide a structured, vendor-neutral standard that improves how enterprises prepare, exchange, and govern document data for AI pipelines. Built on the DocLing toolkit, it is described as 'JSON for documents.' Analysts are broadly supportive of the initiative but raise concerns around governance, accountability, and the risk of forcing new syntax burdens onto users.
125 Impressions