The Paper Authoring Tool began as the Addiction Paper Authoring Tool (Addiction PAT). It was the brainchild of Professor Robert West, who was at that time the Editor-in-Chief of the scientific journal Addiction.

As an editor of a journal and as an addiction scientist, Robert was very much aware of the inefficiency and limitations of study reports, and particularly of randomised trials, which are the key way in which many treatments and prevention programmes are evaluated.

Even with detailed guidance from the EQUATOR consortium (for example, CONSORT), trial reports are typically incomplete and often misleading. Moreover, synthesising evidence from trials to draw conclusions about effectiveness is extremely time-consuming and fraught with the difficulty of deciphering ambiguous and inconsistent use of terms.

The problem came into sharp focus through the Human Behaviour-Change Project (HBCP), the Wellcome-funded collaboration led by Susan Michie from 2017, which set out to use artificial intelligence to read the trials literature and predict the effects of behaviour change interventions. That work exposed just how inadequate published reports were as a source of structured, computer-readable information — and so, in a sense, PAT was the practical answer to a problem the HBCP had made unavoidable.

The idea was that PAT would be an online resource that would prompt users for all the relevant information, store that information in machine-readable form, and provide users with a draft text version of the paper that they could then edit to improve the language and tailor to the needs of whatever journal they were submitting it to.

The first version of the website was built in 2019 as Addiction PAT, hosted at addictionpat.org and supported by the journal Addiction and the Society for the Study of Addiction.

It was a wizard-based application: users worked through a series of linked wizards, each capturing part of the paper, so that information entered in one section could inform the prompts in related sections. This first release was limited to two-arm randomised controlled trials, and was underpinned by the beginnings of an Addiction Ontology (AddictO), which supplied the defined terms used to prompt authors and structure the output.

The tool entered beta testing over 2019–2020. A guide to Addiction PAT appeared in the Qeios journal in January 2020, and a commentary setting out the wider vision was published in Health Psychology in September 2020 as part of the NIH Science of Behavior Change special issue.

The tool was formally introduced to the field in Robert's 2021 paper PAT: an on-line paper authoring tool for writing up randomized controlled trials accompanied by an Addiction Audio podcast.

PAT has been sustained from the outset by many funders. The journal Addiction and the Society for the Study of Addiction supported the original build, with further contributions from Northwell Health, the Wellcome Trust, and Cancer Research UK.

The tool's current development is funded by the National Institutes of Health through APRICOT (Advancing Prevention Research in Cancer through Ontology Tools), the five-year successor to the HBCP running from 2024 to 2029, one of whose explicit aims is to develop tools to facilitate the use of ontologies in behavioural and social sciences.

As its scope broadened beyond addiction, the tool was rebranded from Addiction PAT to the more general Paper Authoring Tool and migrated to a new home at paperauthoringtool.com, with expanded functionality aimed at the behavioural and health sciences as a whole. The current version supports the reporting of randomised controlled trials with any number of arms, as well as pilot and feasibility studies. It allows users to draw labels from the Ontology Lookup Service, links author profiles (including declarations of competing interest) to ORCID for reuse across papers, and continues to export both an editable Word draft and a machine-readable record — realising the original ambition of a tool that serves human readers and automated evidence synthesis at the same time.