To the editor — The COVID-19 pandemic caused an overnight transition to virtual interactions in academia. One non-systematic list noted that 40 out of 59 ecology and evolution conferences in 2021 were online1. In particular, publishing recorded talks online has gained popularity, with the positive intent of increasing accessibility and visibility of scientific events2. Although this has revolutionized how we share findings in ecology and evolution3,4, it also poses novel privacy and ethical issues that can inadvertently risk further impairing inclusion in science5,6. Here, we briefly discuss the opportunities and risks of routinely publishing talks, and highlight considerations for presenters, their hosts and journals.

Publishing talks online gives scientific content a larger and wider reach3,4. It improves accessibility by allowing people to watch the recordings at their convenience (for example time zone) and preferred settings (for example speed)4. It can improve inclusion by allowing access to those who are less able or willing to attend in-person events, which might be owing to attendance costs, visa constraints, caring responsibilities, teaching or fieldwork responsibilities, or environmental concerns. This is especially important for biodiversity conservation in developing countries, where biodiversity loss is highest7 yet access to scientific content, networks and resources is typically lowest8. Publishing talks may promote interdisciplinarity by lowering barriers to access to those who are in other fields9. It can also increase the visibility of researchers’ online profiles to colleagues and the media, and boost public outreach. Such improved reach, however, also accentuates several considerations of concern.

We suggest that three main risks of publishing talks should be actively considered. First, it raises concerns over data privacy and ethics. Photographs and videos of human research subjects, or nonhuman animals in their natural habitats, may unintentionally disclose their identities and whereabouts to a wide audience. This may violate privacy and conservation protocols. Moreover, ethical concerns also apply to the nature of scientific information presented: if the reliability of findings is not obvious to a wider audience because it is not clear whether or not they have been peer reviewed, they could be circulated with disproportionate credibility.

Second, publishing talks can discourage speakers from presenting unpublished work. A main purpose of academic talks is to communicate ongoing research to colleagues, and much of their excitement comes from sharing unpublished work. However, some researchers are discouraged from sharing unpublished work if talks are recorded. This can be because of a real or perceived risk of (1) lost novelty to scientific journals, as there is ambiguity from many journals about what forms of online publication constitute prior publication10; (2) diluting the novelty of a subsequent publication to the media, including lessening the impact of associated images and videos, and/or eliciting premature media coverage; and/or (3) being scooped by other scientists beyond the intended live audience.

Third, aspects of talks or subsequent discussions that speakers might find uncomfortable become open to scrutiny indefinitely. These may include interactions with aggressive questioners; research findings that are subsequently corrected or otherwise improved; or personal aspects of the speaker at that particular time, such as imperfections in the presenting language or inexperience at presenting. Speakers might not wish for this information to be automatically available on record to others, including future employers, and so to define them forever. Similarly, talk attendees might be less likely to ask questions or express opinions if their remarks are recorded. Both speakers and the audience may perceive a risk that anything said can be taken out of context and misinterpreted, reducing the spontaneity of discussion.

The second and third considerations disproportionately affect early-career and already marginalized scientists, and so should be considered alongside the benefits of publishing talks to these groups.

We suggest that speakers should always be given a choice about whether their talk is recorded and published by the host organization. Hosts should not assume consent, and instead present speakers with clear information and guidelines about any intent to record and publish their talk well in advance, such as whether, where and for how long the talk will be online, and who will be able to access it (Fig. 1). If a speaker consents, they can then adjust their talk accordingly — for example, by removing conservation-sensitive or personally identifiable information, or new data that they do not wish to make available yet. Speakers should also make the nature of data clear, including whether they are preliminary and whether they have been peer reviewed. They should also consider including a ‘no tweet’ symbol if they so wish. We suggest that keeping discussion following talks off the record may help to maintain the excitement and spontaneity of sharing ideas. Should organizers decide to record post-talk discussion, they should explicitly communicate this with the speakers and audience. Finally, we urge journals to be explicit about their editorial policies on whether work that has been published in a talk online constitutes prior publication, and on any concerns relating to prior media coverage as a result of such wide exposure.

Fig. 1: Recording and publishing talks.
figure 1

We suggest the following considerations for recording and publishing talks for talk presenters, their hosts and journals.

These recommendations for best practice can be achieved while still running online seminars and conferences, and sharing their benefits for inclusion and carbon reduction3. To strive towards diversity and accessibility, event organizers would ideally focus on maximizing participation in the event itself (whether in person or online), and ensuring it is inclusive to underrepresented groups3,4,6.