Digital Science has launched Papers AI, an AI-native workspace that lets researchers and students write, organise notes and analyse data inside a single environment — with an AI assistant that carries context across the whole project. The press release is dated 4 August 2026 and datelined London, with trade coverage appearing the same day from Research Information, which described the tool as a workspace designed to support the research workflow (Research Information).

The product was built by the team behind Overleaf, the collaborative writing platform that currently reports 20 million users. Papers AI was developed with direct input from the research community, according to Publishing Perspectives (Publishing Perspectives, 6 August 2026).

The workspace handles a broad mix of file types: Word, Markdown, Typst and LaTeX documents sit alongside Jupyter notebooks, CSV data files and Kanban boards. Jupyter cells can be run in place and their outputs reused directly as figures and tables. Every AI-suggested change appears as a reviewable edit — meaning the user can accept, reject or adjust each modification before it becomes part of the document.

One design choice that may matter to cautious institutions: Papers AI can run entirely on local hardware, so no data leaves the user's machine. Users can also plug in an external or local AI model instead of, or alongside, the built-in assistant.

Amye Kenall, Chief Product Officer at Digital Science, and Juan Castro, Principal AI Scientist, are named on the launch. The press release was authored by David Ellis and distributed via the TMX Newsfile wire service, with syndication on Yahoo Finance (Yahoo Finance, 4 August 2026).

The launch is the latest in a dense run of AI-related announcements from Digital Science this year. The company's newsroom lists at least eight product updates or launches since March 2026 — among them Dimensions Research Strategy, an AI analytics platform announced 6 July; Dimensions Citation Check, a research-integrity tool for journal editors announced 16 June; and Altmetric MCP, a Model Context Protocol server that connects AI agents to Altmetric's research-attention tracking data, announced 6 August (Digital Science newsroom). A Model Context Protocol server, for those unfamiliar, is a standardised way for AI tools to pull in external data sources.

For anyone who has watched the academic-tools landscape, the Overleaf pedigree is worth noting. That platform, also a Digital Science property, became the default collaborative writing environment for millions of researchers worldwide — largely by solving a genuine pain point around LaTeX collaboration. Whether Papers AI finds a similar foothold will depend on whether researchers trust an AI assistant inside their workflow, and whether the local-processing option is enough to quiet the privacy concerns that have slowed AI adoption in academia. The tool is available now.