A team of British cybersecurity researchers has published a forensic analysis of Donald Trump’s social media output, identifying systematic patterns of disinformation that echo tactics used by state-backed influence operations. The study, conducted by the University of Cambridge’s Centre for the Study of Existential Risk and the cyber-intelligence firm Graphika, examined over 10,000 posts from Trump’s accounts on Truth Social and X (formerly Twitter) between January 2021 and March 2025. Researchers used natural-language processing and network analysis to map how the former US president’s messages spread through hyper-partisan echo chambers, often amplifying unverified claims before mainstream media could fact-check them.
“We saw a distinct playbook: initial ambiguity, then a coordinated push from loyalist accounts, followed by a narrative shift that made the falsehood seem like a settled fact,” said Dr. Emily Chen, the study’s lead author. The report highlights three key techniques: ‘strategic amplification’ where Trump’s posts are rapidly reshared by bot nets and verified accounts; ‘gaslighting loops’ where the source of a claim is obscured; and ‘prebuttals’ that pre-emptively discredit expected criticism.
The team also noted a worrying convergence with QAnon-style mythology, blurring lines between raw disinformation and meta-narratives about a ‘deep state’. The findings have prompted calls for a legally binding code of conduct for political leaders’ online behaviour, with the UK’s Digital Minister, Sarah Jones, stating that “democracy must not be a laboratory for viral lies.” Trump’s spokespeople dismissed the report as “partisan nonsense from ivory-tower leftists.
” The study arrives as the UK’s Online Safety Bill enters its second reading, adding urgency to debates about exempting ‘journalistic content’ and ‘political speech’ from moderation rules. For the technologist, this is a stark reminder that the real threat is not just the bots or algorithms, but the human vulnerability to narratives that feel true. The team’s dataset and methodology have been released open-source, inviting global peer review.
As one researcher put it: “We’re not trying to ban Trump. We’re trying to make immune systems stronger.
