ABS and open data are more than compatible

Technical solutions can lay the groundwork for ABS without compromising the progress of research.

Dr. Timothée Poisot

TGTTTACTGTTGGTATGGATGTAAAGAGCACGGTATTTTT...

This is how the sequence for GBIF occurrence 4038980442 starts.

Sitting at my desk in Canada, I can access this sequence, and work with it. I know this corresponds to a parasite from the class Monogenea, sampled in Madagascar, and that to use it, I agree to cite the iBOL dataset. Having transparent access to this information makes my work easy.

But it does not make my work fair.

I can shop around GBIF (or other databases) and collect sequence data (digital sequence information, or DSI). I would be reminded of my obligation to give proper credit through citation, but there is nothing that would enforce my obligation to adopt proper Access and Benefit Sharing (ABS) practices. Public sequence information is a public good, and not part of the originating country's genetic resources.

But Madagascar, from where the organism sampled to produce this sequence originates, is clear about the fact that DSI should fall within the scope of the Nagoya protocol: “without genetic resources, there will be no DSI”. And more than a few countries agree.

In an ideal world, there would be a technical solution that allows rapid access to DSI (so that the progress of research is not impeded) while maintaining a country's right to benefit from access to their genetic resources (so that ABS is taken seriously). And so, this begs the question: as open access to DSI and ABS are incompatible, which should we prioritize?

This is the wrong question to ask.

Not only are open access to DSI and ABS perfectly compatible, strong ABS principles are, in fact, more in line with the spirit of open access to research data than the status quo.

ABS is technically compatible with the open data agenda

Making data open is about removing barriers to access. For example, the IUCN red list data are open, in that I can log in to my account, and download them. The GBIF data are open, in that I can log in to my account to request a download, or hit the API to stream them.

Ensuring that ABS principles are respected is not about restraining access to data, but about keeping a record of access for accountability purposes. It is worth considering this as an engineering problem.

In fact, most open databases already (unbeknownst to them) operate the most bare-bones solution required to build proper ABS support: they keep logs. The administrators of any web service can see who downloaded what resource at what time. If this information includes a unique identifier, then the raw material required for ABS already exists.

Requiring authentication to access data is a well-established practice, and there are very few who see this as antithetical to open data principles. Notably, the need for authentication does not preclude automated/scripted access to data, and therefore does not stand in the way of scientific progress.

ABS upholds the spirit of the open data agenda

Although the discourse around open data has often been dominated by individualistic perspectives (“I get to do more research” v. “people get access to my data for free”), any project that seeks truly to collectivize a common good should be thought about in collective terms. Crucially, benefit (to data producers) can only be ensured upon access (by data consumers) if there is a thorough record keeping of who accessed what.

The collection of data supporting enforcement of  ABS is agnostic to the intent behind data access. Although scientific and commercial use of DSI are fundamentally different (and subject to separate obligations), the mechanisms to log data access would apply in all cases. In particular, there is no reason to assume any mechanism of asking permission for use (although some data access context do require this, see e.g. OCAP). Technical solutions to ensure ABS compliance are therefore not resulting in a new barrier between users and data, but are actually a transparent mechanism to ensure that the rights of data producers are protected.

This is a much stronger protection than the current social contract (“please cite this reference”) provides. In fact, transparency surrounding the record of who accessed what resources would be an unprecedented level of traceability. In most cases, this information will never be used, but when access to data result in “utilization of genetic resources along the value chain” (to use the Nagoya protocol parlance), it will allow us to properly discuss how benefits should be shared.

Action on ABS is urgent

Stalled discussions on the WHO Pandemic Agreement have highlighted how contentious the idea of Access and Benefit Sharing for Digital Sequence Information is: it is apparently worth delaying our global capacity to prepare for, and respond to, the next pandemic.

But we can start building tools that allow a clear assessment of ABS needs and responsibilities, even in the absence of global agreement on mechanisms about the use of this information. This is because, in practice, working under a strict ABS regime (all access to data is logged and attributed to an individual) and working under the status quo (almost all access to data is logged but not always attributable to an individual) makes no difference on the daily work.

A lot of the considerations holding back Pandemic Agreement negotiations are about mechanisms that would happen post establishment of ABS technical solutions. The lack of consensus on how the data regarding access to DSI should be used in the future is not a strong argument to delay the implementation of technical solutions that support the collection of such data in the short term. As it is, the scientific and legal communities are trying to solve three problems at once: the agreed-upon implementation of ABS, the governance of data enabling monitoring of ABS, and the mechanisms to act on the data collected this way. As we can make realistic (and immediate) progress on the first two, by adopting traceable access to data, with essentially no disruption to scientific work, it feels like a low-hanging fruit that may, in term, contribute to a more equitable landscape.