Social media reshapes the way people seek health answers, argues Dr. Deborah Cohen in her book Bad Influence (Oneworld, 2026), which has been shortlisted for the 2026 Royal Society Trivedi Science Book Prize. Cohen documents how platforms amplify peer stories and support networks while simultaneously creating fertile ground for commercial promotion, misinformation and fast-paced trends. Central to her account is the role of recommendation systems: algorithms that prioritise engagement can make certain treatments and conditions more visible, shaping public perception of what is common, urgent or desirable.
Her reporting highlights a persistent mismatch between expectations shaped online and the limits of clinical practice. Social feeds often offer neat narratives and apparent certainty; formal services such as the NHS deliver care within clinical constraints and uncertainty. That contrast, Cohen says, can drive people toward quick fixes and commercially backed solutions rather than evidence-based care. She raises the prospect that rising use of large language models will reconfigure how people interpret symptoms and treatment options, underlining the urgent need for improved health literacy.
Cohen also examines the dynamics of influence: some health professionals and content creators cultivate intimate, trust-like relationships with followers and use that authority to promote products or services. Regulators such as the GMC face new questions about professional conduct in public fora. The book documents how topic cycles—attention around conditions like ADHD or menopause—can be amplified because engagement equals revenue, not necessarily because of any change in underlying prevalence or evidence.
The practical implications span public communication, platform design and clinical practice. Cohen suggests that official health messaging may need to adopt more engaging formats to compete with short-form storytelling, while policymakers must consider algorithmic harms and incentives. Emerging tools such as LLMs could support lay understanding of medical jargon, but they also carry risks of error and flattering, reassuring outputs. Recent survey findings cited in her work indicate that a large share of younger adults already use social media for health information, underscoring that any response must bridge digital behaviour, regulation and basic scepticism in order to reduce harm and improve how people find trustworthy care.
