Many ideas focused on improving scholarly communications are borrowed from adjacent industries, where there is some overlap in concepts or procedures that lend themselves to more easily envisaging how new ways of working may be adopted. Think the use of Agile project management processes in publishing, or the increasing adoption of trust markers in research integrity issues. Ideas that make sense in another context and can be mapped to our own area have a better chance of being adopted to improve outcomes of immediate concern.
One concept that is used occasionally – if informally – is that of due diligence from legal and commercial fields. The term is used somewhat loosely, and one can imagine a research supervisor or publishing head inquiring of their staff members if they had ‘used due diligence’ in checking the research literature or an author’s background. But what if it was more tightly defined and had more of a codification? Could it help the research integrity crisis we seem to be suffering?
Safety net
Due diligence can be defined in a number of ways depending on the specific context. In law, it is defined according to Merriam-Webster as “the care that a reasonable person exercises to avoid harm to other persons or their property”, whereas in business the definition covers “research and analysis of a company or organization done in preparation for a business transaction”. The focus is subtly different for each definition, with the emphasis on personal damage in the legal sphere as opposed to risk management in commercial areas.
What can we learn from this in scholarly communications? Some people might use due diligence already in an ad hoc sense, but there would be greater utility in having some form of framework and clear definition for our specific context. We do have frameworks that make some form of due diligence implicit – for example, the recently updated COPE its Code of Conduct for members as well as its Core Principles. The latter is particularly interesting as it positions ‘Accountability’ as the first principle, citing the need for actors to “take responsibility for policies, procedures, decisions, business practices and other activities related to research and scholarly publishing”. The key aspect of this is perhaps what is not included, ie. An explicit reference to the (responsible) use of AI and its lack of agency.
Use cases
Using the idea of due diligence, the industry could perhaps take COPE’s example one step further, with each agency involved making themselves more accountable for the research publication process.
- At the start, researchers can register their idea and state they have checked the literature to note any similar studies
- Where funders are involved, they can publicly state the support they have given as well as the expectations they have for any subsequent publications
- For universities, they will have an increasingly important role to play as the provider of AI tools – with this also comes greater responsibility to both support and monitor their faculty’s research conduct and outputs
- For publishers and editors, they can go beyond the adoption of AI surveillance tools and COPE membership by running further checks on authors’ prior research and public statements made by other agencies listed above.
If the common thread for due diligence in other fields is preparation, then this also covers the need for each agency not only to lay the groundwork for what they are doing, but to be doing so transparently. Like the principle of blockchain, each agency mutually supports the others, enabling the trustworthy sharing of information. This does happen now after a fashion, but the crucial difference is that it relies on convention rather than an explicit code. COPE’s changes are a positive start on the road to better managed integrity, but we can do more with a little extra (due) diligence.