# Workers Retroactively Add AI Language to LinkedIn Profiles > Source: > Published: 2026-08-24 13:05:10+00:00 # Workers Retroactively Add AI Language to LinkedIn Profiles A National Bureau of Economic Research working paper finds that 19.7% of established U.S. LinkedIn users retroactively edited a past job title or description between 2020 and 2026. Using monthly Revelio Labs data covering 29.4 million profiles, the researchers documented a sharp post-2022 rise in AI-related language, alongside declines in remote-work and DEI terminology. A National Bureau of Economic Research working paper finds that workers are increasingly revising descriptions of jobs they have already left to include AI-related terms, while references to remote work and DEI have declined. The July 2026 paper, "Time Travel on Professional Profiles," analyzes monthly Revelio Labs snapshots of 29.4 million U.S. LinkedIn profiles collected from August 2020 through January 2026. The researchers define a "time-travel" edit as a change to the title or description of a role made more than 60 days after a worker left that position. According to the paper, **19.7% of established U.S. LinkedIn users** made at least one such retrospective edit during 2020-2026. Revelio Labs reports that the median edit occurred **4.5 years** after the position ended. ### AI terms rise after 2022 The NBER paper documents sharp increases after 2022 in AI-related language added to older job descriptions, including references to AI, GPT, LLMs, and artificial intelligence. Business Insider reports that retroactive additions of those terms rose more than sixfold after late 2022. The paper's authors write that retroactive profile revisions can reveal workers' beliefs about labor-market demand. That is an interpretation of the observed text changes, rather than a verification that workers acquired new skills or performed newly claimed responsibilities in the earlier role. Gideon Slocum Moore, a Stanford researcher and paper co-author, illustrated the distinction in comments reported by Fast Company and NDTV Profit: a worker who previously listed a recommendation-systems role as "machine learning engineer" could later change it to "machine learning and AI engineer" after AI terms became more valuable in the job market. Because the underlying role had already ended, Moore said, the change concerns how that experience is presented. ### Remote-work and DEI language falls The study also found fewer additions of remote-work language, such as "remote," "hybrid," and "WFH," to prior roles. Business Insider reports that by the end of 2025, users were adding those terms about as frequently as they were deleting them, a change that coincided with return-to-office mandates. Separately, the researchers observed a sharp early-2025 decline in additions of terms including diversity, equity, inclusion, and DEI. Revelio Labs links the timing to President Donald Trump's January 2025 executive order ending federal DEI programs. The research documents a temporal association between the policy environment and profile-editing behavior; it does not establish that the order caused any individual edit. ### A job-transition indicator, with data-quality implications Retroactive editing clustered around employment changes. Revelio Labs reports that editing activity starts rising about six months before a job transition, peaks around the move, and declines afterwards. The NBER paper similarly finds that workers making such edits are more likely to change employers than users who edit later. The practice was more prevalent in some groups. Fast Company reports a 38.1% retrospective-editing rate among users under 30, 31.6% among workers in technology and information, and 29.3% among MBA holders. It also reports especially high editing rates in AI leadership, design, qualitative research, and UX research roles. For labor-market researchers and teams using professional-profile data, the paper raises a measurement issue: historical profiles are not fixed records. Skill-demand analyses that treat current profile text as a contemporaneous account of past work can incorporate later revisions. Conversely, the timing and wording of revisions may provide a useful, though indirect, signal of how job seekers respond to changing terminology in hiring markets. The paper also identifies a post-ChatGPT increase in LLM-associated writing markers in profile edits. According to the NBER researchers and Revelio Labs, these markers were more common among workers without college degrees and among MBA graduates from lower-ranked programs than among workers with more education or MBAs from higher-ranked programs. The markers indicate language patterns associated with LLM-assisted writing, not confirmed use of a particular AI system. ## Key Points - 1The study finds 19.7% of established U.S. LinkedIn users retrospectively edited past roles, making profile history a mutable labor-market dataset. - 2AI-related terminology rose sharply in old job descriptions after 2022, while remote-work and DEI additions declined, according to the researchers. - 3For workforce analytics, historical skill counts derived from current profiles can reflect later marketing edits rather than contemporaneous job responsibilities. ## Scoring Rationale The research is notable for practitioners using LinkedIn and other professional-profile data for skill-demand, hiring, or workforce analyses. It provides large-scale evidence that profile text changes over time, but it does not introduce a new model, tool, or regulatory requirement. ## Sources Primary source and supporting public references used for this report. ## View 5 more sources [Revisionist (Employment) Historyreveliolabs.com](https://www.reveliolabs.com/news/ai-and-work/revisionist-employment-history)[LinkedIn users are 'time traveling' by adding AI keywords ...fastcompany.com](https://www.fastcompany.com/91589906/linkedin-users-are-time-traveling-by-adding-ai-keywords-to-past-jobs-on-their-profiles)[What LinkedIn 'time traveling' means for employee retentionbizjournals.com](https://www.bizjournals.com/bizjournals/news/2026/08/13/linkedin-profile-ai-llm-remote-work-dei-jobs.html)[Workers are rewriting their LinkedIn histories to add AI skills. Here's what they're dropping.businessinsider.com](https://www.businessinsider.com/linkedin-users-add-ai-skills-jobs-drop-remote-work-terms-2026-8)['LinkedIn Time Travel': 20% Users Retroactively Edit Past ...ndtvprofit.com](https://www.ndtvprofit.com/technology/linkedin-time-travel-20-users-retroactively-edit-past-jobs-and-ai-is-a-top-addition-11911375) Practice with real Social Media data 90 SQL & Python problems · 15 industry datasets 250 free problems · No credit card [See all Social Media problems](/problems/datasets/social)