The Ethics of Authorship Authors: Lauren Venning, Emma Keelan Published: 2026 In modern postgraduate medical training, academic authorship has evolved from a desirable addition to a ‘musthave’ credential. Somewhat concerningly, in several UK specialty training application processes, the strength of an academic CV now outweighs direct evidence of a doctor’s clinical skills or knowledge. As a result, for resident doctors seeking to progress into specialty training, posters and publications have become the necessary currency for success. While the development of academic writing and presentation skills offers genuine educational value, the current emphasis on publication accumulation “for points” has generated significant tensions. Residents may feel conflicted between rapidly producing low-impact outputs to satisfy application requirements and pursuing meaningful research that contributes substantively to medical knowledge. This pressure is further compounded by the financial burden of conference attendance and publication fees, often incurred at the expense of investment in essential clinical examinations. Such examinations not only enhance patient safety but also play a critical role in the development of competent future clinicians. Navigating the already complex and opaque landscape of medical authorship presents an additional challenge for resident doctors, a difficulty that has been further intensified by the rapid advancement of artificial intelligence (AI). This commentary examines the prioritisation of academic competence within the medical specialty application process, focusing on three intersecting issues: (1) the privileging of publication quantity over academic quality and clinical competence; (2) the persistence of gift authorship and consultant pressure; and (3) the emerging ethical and practical challenges of AI-assisted authorship. Publishing for Points The current UK Physician Higher Specialty Training Recruitment (PHST) process allocates interview selection based on a self-assessed score across clinical, academic, and leadership domains. Overseen by the Royal College of Physicians through the Physician Specialty Recruitment Office (PSRO), this system is intended to standardise selection into Internal Medicine Training (IMT). In practice, however, publication records exert a disproportionate influence on shortlisting. A first-authored original research paper in a peer-reviewed journal attracts the highest academic score, yet applicants may accrue equivalent (or greater) points through multiple case reports, abstracts, or letters. While such outputs have educational value, they may carry substantially less scientific weight than sustained academic training, such as higher research degrees or long-term projects. The consequences of this system are twofold. First, resident doctors are acutely aware that without publications, progression may be unlikely regardless of clinical excellence, with many capable clinicians being excluded from interviews solely on the basis of portfolio scoring. Second, the structure actively incentivises a “publish for points” culture, in which scholarly activity becomes instrumental rather than intrinsically academic. As Stehlik et al. observed in their analysis of mandatory research requirements in Australia and New Zealand, such systems can lead to “unintended consequences”, including low-quality research or poorly planned and executed projects.”1 Time spent producing minimally contributory manuscripts detracts from clinical training and may expose trainees to superficial collaborations, poor academic practice, and ethical vulnerabilities, all in service of portfolio optimisation rather than scientific or clinical value. What Constitutes Authorship? The International Committee of Medical Journal Editors (ICMJE) defines authorship as requiring substantial contribution across four domains: conception or design of the work; acquisition, analysis, or interpretation of data; drafting or critically revising the manuscript; and final approval of the published version.2 Authorship also entails accountability for all aspects of the work. However, under pressure to secure first-author publications, resident doctors often encounter ethically problematic situations. These may include being expected to produce manuscripts primarily for senior colleagues, sometimes with limited supervision or educational benefit. Such demands add to workload strain and increase burnout risk, as trainees attempt to balance clinical duties, examinations, and academic output. Medicine’s hierarchical culture further complicates matters. Gift authorship, the inclusion of individuals who do not meet recognised authorship criteria, remains widespread despite being unethical.3 A 2020 analysis found that 41% of Cochrane reviews contained gift authorship, even though the practice constitutes intellectual corruption.3,4 Alarmingly, 15% of first authors were unaware of ICMJE authorship criteria, highlighting the lack of formal education in this area.4 Gift authorship commonly occurs in an “upward” direction, where junior doctors feel compelled to include senior clinicians due to perceived obligation, implicit pressure, or fear of repercussions. In some cases, trainees may deliberately include senior figures to lend credibility and improve publication prospects. Lateral gift authorship between peers and “gift-down” authorship from seniors to juniors may also occur, often to inflate publication counts. While the prevalence of ‘lateral’ authorship is difficult to quantify, the allocation of application points for roles such as joint first author, corresponding author, or co-author may inadvertently facilitate this form of honorary authorship. This structure risks normalising reciprocal arrangements among resident doctors, whereby collaborative authorship is pursued primarily for mutual portfolio benefit rather than genuine scholarly contribution. Regardless of intent, such practices undermine transparency, accountability, and fairness. They breach publication ethics, contravene the General Medical Council’s principles of honesty, and distort responsibility for the scientific record.5,6 Most concerningly, they disadvantage those most vulnerable; typically women, international doctors, non-substantive/locum staff and junior colleagues, who undertake the majority of the scholarly labour while bearing the greatest professional risk. To be clear, clinical involvement in patient care alone does not justify authorship.2 Similarly, supervision without substantive intellectual contribution does not meet ICMJE criteria. Addressing this requires explicit education on authorship ethics and visible leadership from consultants, including declining authorship when criteria are not met and resisting institutional pressures that equate academic worth with publication count. AI and Authorship The rise of artificial intelligence introduces a further challenge to authorship integrity within an already strained academic system. Tools such as ChatGPT and GPT-4 can generate ideas, draft manuscripts, restructure text, and summarise existing literature.7 When used transparently, such tools may help level the playing field for time-pressured clinicians, non-native English speakers, and those with additional caring responsibilities. However, undisclosed AI use represents another form of authorship without accountability.8 Detection of AI-generated content remains challenging. Dunn et al. demonstrated that manuscripts produced by ChatGPT were largely indistinguishable from human-written text.9 This has direct implications for fairness in trainee selection, as applicants who rely heavily on AI may gain an artificial advantage over peers who invest significant time and effort in genuine research and writing. While leading journals such as Nature and JAMA now mandate disclosure of AI use, many trainee-led or lower-tier journals lack clear policies, creating ethical inconsistency across the academic landscape.10 Undisclosed AI involvement also weakens accountability. Large language models may fabricate references or propagate inaccuracies, which can then enter the scientific record under human authorship.10,11 Given that a substantial proportion of trainee publications are already low impact, with 59% failing to receive any citations, the risk of further diluting research quality is considerable.12 As per the International Committee of Medical Journal Editors, AI cannot be listed as an author because it cannot assume responsibility or accountability.2 The unresolved question, therefore, is not whether AI can be used, but the extent to which its use is acceptable and how transparently it should be acknowledged across all levels of medical publishing. Potential Solutions Education is central to addressing many of the issues outlined above; however, it must be accompanied by a cultural shift driven from the top down. Formal training in authorship ethics should be embedded within both undergraduate and postgraduate medical curricula. Early and repeated education on ICMJE authorship criteria, gift authorship, and academic accountability would better equip trainees to navigate authorship decisions with confidence and ethical awareness. Crucially, such training must also extend to consultants and supervisors. Explicitly addressing gift authorship as an ethical issue reinforces shared responsibility for academic integrity and encourages those in positions of authority to model appropriate behaviour. This would help trainees to challenge inappropriate practices without fear of reprisal. Recognising and rewarding supervisors for high-quality mentorship rather than publication volume alone may further help shift the culture away from performative authorship towards meaningful scholarly engagement. In addition, the routine use of written authorship agreements at the outset of projects, including posters and written submissions, should be normalised. Early clarification of roles, expectations, and authorship criteria can reduce misunderstandings, mitigate hierarchical pressure, and protect junior researchers. While such agreements cannot eliminate power imbalances entirely, they do provide a framework for transparency and accountability that is currently absent from many trainee-led projects. That being said, addressing the role of artificial intelligence in authorship presents a more complex challenge, largely due to the absence of comprehensive guidance. Although the ICMJE states that AI use should be documented and that chatbots cannot be listed as authors, there is currently no mechanism to regulate or define acceptable levels of AI involvement in academic work.2 As such, it is imperative that selection bodies, journals, and training programmes develop unified standards for AI disclosure and responsible use. Education on the capabilities and limitations of AI tools, including the risks of fabrication, bias, and over-reliance, should form a core component of academic skills training at undergraduate and postgraduate level. Used transparently and ethically, AI has the potential to enhance scholarly communication; used covertly, it risks further eroding trust in academic outputs. Conclusion Authorship disputes are not new in medicine; however, the growing emphasis on publication metrics and the integration of artificial intelligence have intensified longstanding ethical challenges. Whether credit is granted to a consultant who did not meaningfully contribute, or derived from reliance on an algorithm incapable of assuming responsibility, the underlying issue remains the same: authorship without accountability. These concerns underscore the need to re-evaluate specialty recruitment systems that prioritise academic outputs over demonstrated clinical competence. Trainees who wish to engage in research should be supported through protected time, appropriate supervision, and structured training, rather than incentivised to produce minimal publications for application points. A shift in emphasis from publication quantity to research quality and integrity is urgently required. As Douglas G. Altman cautioned, “We need less research, better research, and research done for the right reasons.”13 Reaffirming this principle within postgraduate training is essential if authorship is to remain a meaningful marker of merit rather than a hollow credential.