TikTok user reports rarely triggered action on UK drug-selling accounts

A prospective audit of TikTok’s user-reporting system found that the platform issued an explicit “no violation” decision for 95.2% of reports concerning UK-facing accounts that researchers identified as offering or facilitating the supply of illegal drugs. The study, led by the University of Bath with researchers from University College London and the University of Manchester, also found that 83.5% of the reported accounts remained publicly accessible after 14 days. Researchers described the experiment as an independent, real-time test of what happens when an ordinary user directly reports apparently drug-selling content through the standard in-app mechanism, rather than an assessment of the platform’s broader proactive moderation performance.

Over five months, from February to July 2026, the team searched for TikTok accounts aimed at UK users and displaying a drug-related cue together with an offer or route to supply. The researchers selected 455 accounts that met those criteria, including accounts referring to cocaine, synthetic cannabinoids commonly known as spice, and controlled medicines. Accounts qualified only when they combined evidence of drug-related content with UK targeting and an apparent supply route. This screening process meant that the study focused on content that the authors argue represented a relatively straightforward test of TikTok’s reactive reporting pathway, although the protocol involved researcher identification of accounts rather than a representative sample randomly drawn from the platform.

For each account, the researchers used TikTok’s ordinary in-app reporting function, recording the platform’s response and subsequently checking whether the account remained accessible. The “no violation” outcome accounted for 95.2% of reports, while 83.5% of accounts were still publicly accessible 14 days later. The researchers reported the same pattern across different drug categories and forms of promotion. They also observed that most changes in accessibility occurred within the first two days, with limited removal thereafter. Every decision classified as “no violation” was returned exactly 30.5 minutes after the report, and accounts that survived that initial interval were unlikely to be removed later.

TikTok’s Community Guidelines prohibit the trade, marketing, and promotion of illegal drugs. The study also examined the policy context created by the Online Safety Act 2023, which classifies drug-supply offences as priority offences. Since 17 March 2025, regulated user-to-user services have been required to take proportionate measures to prevent users from encountering priority illegal content and to maintain systems intended to minimise how long such content remains available and remove it swiftly once the service becomes aware of it. Supplying or offering to supply a controlled drug is also a criminal offence under the Misuse of Drugs Act 1971. The audit did not, however, determine whether TikTok’s conduct constituted legal non-compliance; its results instead raise questions about the effectiveness of the systems and processes platforms operate in response to direct user reports.

Study leader Chris Pudney, a professor in Bath’s Department of Life Sciences, said the results expose a weakness in the reactive reporting pathway tested. “This is a straightforward test of a reporting system,” he said. “These were accounts we identified because they displayed drug-related cues, targeted UK users and contained an explicit offer or route to supply. Yet 95% of our reports came back ‘no violation’.” He contrasted those outcomes with platforms’ published proactive-removal statistics, arguing that high numbers of automatic removals do not show whether a service responds when a member of the public points to apparent criminal activity. “Our findings show a serious weakness in the reactive reporting pathway we tested,” Pudney said.

The study’s focus on ordinary reporting is important because it addressed a different question from the one answered by aggregate enforcement figures. TikTok reports removing large quantities of violating content through its own proactive moderation systems, but the Bath-led experiment asked what occurred after a user submitted a specific report using the tool available to any member of the public. The researchers said their findings should not be interpreted as showing that every instance of drug-selling content escapes detection. Nor was the experiment designed to estimate the prevalence of illegal drug offers across TikTok. Instead, it assessed the platform’s documented or practical response to a set of visibly suspicious accounts submitted through the user-reporting channel.

Automation may have formed part of that response, although the researchers could not determine whether individual decisions were made by automated systems, human moderators, or a combination of both. Pudney noted that the identical 30.5-minute timing for every “no violation” response suggests that at least some automation was likely to be involved. The team did not test the internal operation of TikTok’s moderation systems, and it could not establish the reasons for particular decisions. The five-month observation window and 14-day follow-up do show that, among the selected accounts, later removal was uncommon once the early stage had passed. The design therefore offers evidence about real-world performance under a defined protocol, rather than a complete account of the platform’s enforcement activity or technical architecture.

The research team said the ease with which qualifying accounts could be found was another concern, while stopping short of estimating platform-wide prevalence. Hundreds of eligible accounts were identified through ordinary searches using widely known drug-related terms and sales-related language. Because the sample was assembled through researcher searches and then tested through reports, the results do not provide a denominator for all UK-facing drug-selling accounts, nor do they show how many similar accounts existed outside the sample. The study can establish that the tested reporting pathway returned “no violation” for most submitted reports and that most sampled accounts remained accessible after two weeks. It cannot establish the total volume of drug-selling content, the proportion of all prohibited content that remains online, or whether the pattern holds across other platforms, offences, jurisdictions, or reporting methods.

Pudney said the findings raise an obvious question for Ofcom about whether the systems platforms have in place work effectively in practice when users report suspected illegal content. “The Online Safety Act treats offering illegal drugs for sale as a priority offence,” he said. “These findings do not by themselves establish legal non-compliance, but they do raise an obvious question for Ofcom: are the systems and processes platforms have in place actually effective when users report this kind of content?” He and the team are calling for independent, offence-specific audits to evaluate platform responses. Such audits, they argue, should examine what happens after direct reports rather than relying only on platform-generated summaries of proactive removals. The peer-reviewed study, “A prospective audit of TikTok’s response to UK drug-selling accounts,” is scheduled for publication in the International Journal of Drug Policy on 2 November 2026.

Subject of Research: An audit of TikTok’s user-reporting responses to UK-facing accounts advertising and offering illegal drugs for sale.

Article Title: TikTok returns ‘no violation’ decision on 95% of reports of UK-facing accounts advertising drugs, study finds

Article References: TikTok returns ‘no violation’ decision on 95% of reports of UK-facing accounts advertising drugs, study finds. Original publication (Bibliographic details unavailable.)

Image Credits: AI Generated

DOI: Not provided

Keywords: TikTok, drug selling, Online Safety Act, content moderation, user reports, platform enforcement, UK regulation, illegal drugs, Ofcom, social media, returns, violation