Do ministers' written answers sound different since ChatGPT? 42,082 parliamentary answers and 17,292 written statements, 2015 to 2026

SGR-022, conducted 2026-10-22. Analyst: Jack Stovell, founder, ScriptGrain.

Question: Within the same UK government departments, do written parliamentary answers and written ministerial statements given after ChatGPT use more of the words and habits associated with AI writing than those given before it?

Summary

Background. Ministers answer thousands of written parliamentary questions a year, and the answers are drafted by civil servants under pressure to be clear, accurate and quick. That is where people suspect AI tools may have crept in. Whether the writing changed after ChatGPT can be measured, because every answer and every written statement is public, dated and attributed to a department.

What we did. We fixed six hypotheses before the full measurement, copying the design of our studies of UK brands (SGR-011) and GOV.UK press releases (SGR-020). We listed every answered written question tabled from October 2014 and drew a fixed-seed sample of 60,000 from 668,536, stratified by year and department, then fetched every sampled answer and every written statement through Parliament's public API. 42,082 answers and 17,292 statements of 50 words or more, given between January 2015 and September 2026, were measured in code. Each department (renamed and merged departments joined into their lineage) was compared only with itself: 18 of the 19 departments qualified, through 35 department-and-house pairs with at least 10 answers in both 2019 to November 2022 and 2024 onwards.

What we found. AI-associated words rose 26% within the same departments' answers (0.54 to 0.67 per 1,000 words). The words that moved were AI-era marketing words, the quieter LLM-favoured words and the publicly mocked ChatGPT words; the 2024 tells such as "delve" did not move beyond chance. Em dashes did not rise beyond chance (0.54 to 0.47 per 1,000 words). AI-style sentence shapes did not rise beyond chance (0.59 to 0.67 per 1,000 words). Average sentence length stayed within 1.5 words (25.0 to 24.9). By calendar year the words peaked in 2025 at 0.84 per 1,000 words, against 0.61 in 2019 to 2022, and stood at 0.76 in 2026. In written ministerial statements, the secondary series, the same words rose (+37%) and em dashes rose (+46%). In the Conservative months of 2024 alone the change was within chance, so this study cannot separate the rise from the change of government. The same measure moved brand blogs and press releases +40% (SGR-011, 37 brands); GOV.UK press releases and news stories +22% (SGR-020, 20 departments); in these answers the change was +26%.

What it means. Parliamentary answers shifted in vocabulary after ChatGPT, and nothing here shows a break from the trend before 2022. The questions MPs and peers write themselves are a comparison: their AI-associated words changed by +15%, within chance. A change in a word list describes style; it does not show who, or what, wrote any answer.

Key numbers

• UK government departments' written parliamentary answers used 26% more AI-associated words in 2024 to 2026 than the same departments' answers from 2019 to November 2022 (0.54 to 0.67 per 1,000 words; ScriptGrain, 2026; 18 departments, pre-registered).
• Em dashes in written parliamentary answers went from 0.54 to 0.47 per 1,000 words between 2019 to 2022 and 2024 to 2026 within the same 18 departments (change −0.06, 95% interval −0.15 to 0.02).
• AI-associated words in written parliamentary answers peaked in 2025 at 0.84 per 1,000 words, against 0.61 in 2019 to 2022 and 0.48 in 2015.
• The study measured 42,082 written answers and 17,292 written ministerial statements (16,209,936 words in all) from 19 departments and other answering bodies, given between January 2015 and September 2026.
• In written ministerial statements, AI-associated words went from 0.73 to 1.00 per 1,000 words within the same 16 departments (+37%).
• Within-organisation change in AI-associated words after ChatGPT: +26% in parliamentary answers (18 departments), +22% in GOV.UK press writing (20 departments), +40% in UK brands (37 brands).

Sample

Every written question answered between January 2015 and September 2026 that Parliament's Questions and Statements API listed (668,536 answered, non-withdrawn, non-holding answers), sampled with a fixed seed to 60,000 (up to 5,000 a year, at least 200 per department a year where the department has that many), plus every written ministerial statement. 42,082 answers and 17,292 statements were measured after exclusions, from 19 departments and other bodies. The confirmatory tests use the departments only; 35 department-and-house pairs qualified.

• Written parliamentary question answered between 1 January 2015 and 30 September 2026, or written ministerial statement made in that window
• Not withdrawn and not a holding answer
• At least 50 words of text after tables were removed
• At least 10 answers before and 10 after in the same department and house, for the confirmatory tests

Method

Corpus: written questions were listed through the UK Parliament Written Questions and Statements API by week of tabling, at no more than three requests a second with an identified user agent; a fixed-seed sample was drawn by year and department and its full answer text fetched; every written statement was listed by month and fetched. Source: UK Parliament, Open Parliament Licence.

Sampling: for each year, up to 5,000 answers, with at least 200 per department where available and the rest shared in proportion to size; within each year and department, answers ranked by a hash of the question ID and the seed, first ones taken. Departments' means are weighted by the inverse of the chance of selection.

Departments: each answer's answering body. Nineteen departments are analysed, with renamed, merged and split departments joined into one lineage (for example the Department of Health with the Department of Health and Social Care). Other bodies are analysed separately as exploratory.

Text measured: the answer (or statement) reduced to plain text, with tables removed; the question and title are not part of the answer's text. Texts under 50 words, holding answers, exact duplicates within a department and texts where more than 40% of the words sit in short lines or more than half are numerals were excluded.

Measures: computed in code by ScriptGrain's own published instruments, the same as in the brand series and SGR-020. No language model read or scored any text.

Pre-registration: the hypotheses, eras, sampling design, exclusions and tests were committed to the public repository before the full measurement (docs/research/sgr-022-preregistration.md). The run was done once, on the day before publication.

Tests: within-department change in department-weighted means, with 95% intervals from 2,000 bootstrap resamples of departments (fixed seed). H2 compares the per-year rate of change from 2015-18 to 2019-22 with that from 2019-22 to 2024-26, for cells with all three eras. H6 repeats H1 using only 1 January to 4 July 2024. Analyses beyond H1 to H6 are labelled exploratory.

Variables measured

• AI-associated words per 1,000 words (and each word list)
• AI sentence shapes per 1,000 words
• Em dashes per 1,000 words
• Average sentence length
• Contractions per 1,000 words
• "You" share

What each measure means

Before and after

Before: answers given from January 2019 to November 2022 (ChatGPT launched on 30 November 2022). After: January 2024 to September 2026. December 2022 to December 2023 is left out of the main test as a transition year and reported separately. The 2015 to 2018 answers are used only to test whether the change is a continuing trend.

Within-department change

For each department and house (Commons or Lords), the average after an answer minus the average before; then averaged over departments, each counting once. A department is compared only with itself, so a shift in which departments answer more cannot create a trend. Averages are weighted by the inverse of each answer's chance of being sampled, so they estimate the department's whole output. Renamed and merged departments are joined into one lineage.

AI-associated words

Uses per 1,000 words of three published word lists, each occurrence counted once: the 2024 words studies of scientific writing found language models over-use ("innovative", "utilized", "advancements", "pivotal", "facilitates", "firstly", "tackle", "showcasing"…), words publicly mocked as ChatGPT tells, and AI-era marketing words ("seamless", "elevate", "unlock", "empower", "streamline", "supercharge", "unleash", "effortless"…). The same measure as SGR-007 and SGR-011. A separate list of quieter words ("crucial", "comprehensive", "insights", "additionally", "potential", "significant"…) is reported on its own.

AI sentence shapes

A fixed scan in code for 'not X but Y' balancing, reflexive lists of three, chains of semicolons and stock phrases, per 1,000 words.

Dates

The date Parliament records as the day the question was answered (for a statement, the day it was made). Corrected answers keep their original date and are marked; a check without them is reported.

95% interval

From 2,000 bootstrap resamples of departments. A hypothesis counts as supported only if the interval excludes zero in the predicted direction (for H5, if the 90% interval sits inside ±1.5 words).

Findings

• H1 (supported): AI-associated words rose within the same departments' answers (+0.14 (95% interval 0.03 to 0.25) per 1,000 words, +26%): From 0.54 to 0.67 per 1,000 words across 18 departments. By word list: AI-era marketing words +77% (interval clears zero); the quieter LLM-favoured words +13% (interval clears zero); the 2024 tells such as "delve" −13% (within chance); the publicly mocked ChatGPT words +169% (from a small base of 0.013 per 1,000 words; interval clears zero). Leaving out any one department, the change stays between +0.11 and +0.16.
• H2 (not supported): the change after ChatGPT was not clearly different from the trend before (0.030 per 1,000 words a year after, against 0.005 a year from 2015-18 to 2019-22): For the 18 departments with answers in all three eras, the difference in yearly change (+0.024, 95% interval −0.027 to 0.069) includes zero. These words were already moving in books and the web before ChatGPT, as SGR-007 found.
• H3 (not supported): em dashes did not rise (0.54 to 0.47 per 1,000 words; change −0.06 (95% interval −0.15 to 0.02)): Em dashes and spaced en dashes are counted together. Despite the 'ChatGPT dash' reputation, these departments used dashes at a rate within chance of the earlier one.
• H4 (not supported): AI sentence shapes did not rise (0.59 to 0.67 per 1,000 words; change +0.08 (95% interval −0.03 to 0.19)): 'Not X but Y' constructions, reflexive lists of three, chains of semicolons and stock phrases, per 1,000 words.
• H5 (supported): sentence length stayed within 1.5 words (25.0 to 24.9 words; change −0.13 (95% interval −0.48 to 0.20)): The 90% interval (−0.41 to 0.15) sits inside ±1.5 words.
• H6 (not supported): the Conservative months of 2024 alone do not show a clear rise (+0.11 (95% interval −0.09 to 0.34) per 1,000 words (1 January to 4 July 2024 against 2019 to 2022)): 18 departments qualified. Parliament was dissolved from 30 May to 4 July 2024, and this window holds only about five months of answers. This study cannot say whether the change of government or the time since ChatGPT explains the shift.
• Written ministerial statements (same six tests, secondary) (AI-associated words +0.27 (95% interval 0.13 to 0.41) per 1,000 words (16 departments)): Em dashes +0.33 (95% interval 0.13 to 0.56); AI sentence shapes +0.02 (95% interval −0.14 to 0.16); average sentence length −0.13 (95% interval −0.56 to 0.36). Statements are a census, fewer in number than answers, so intervals are wider.
• The questions MPs and peers write (comparison series) (AI-associated words +0.04 (95% interval −0.06 to 0.14) per 1,000 words): Answers' change minus questions' change in the same departments: +0.10, 95% interval −0.04 to 0.24 (exploratory). Questions are written by MPs, peers and their staff, not by civil servants, so they show what a change in the language of Parliament itself would look like.
• Robustness (H1 with the transition year counted as after +0.12 (95% interval 0.02 to 0.22); answers of 150+ words +0.10 (95% interval −0.16 to 0.35)): Unweighted means +0.13, 95% interval 0.02 to 0.25; pooled counts over words +0.17, 95% interval 0.02 to 0.32; without corrected answers +0.14, 95% interval 0.02 to 0.25; Commons +0.23, 95% interval 0.13 to 0.32; Lords +0.03, 95% interval −0.15 to 0.20. Labour months (from 5 July 2024) against 2019 to 2022: +0.15, 95% interval 0.03 to 0.28.

What this study does not show

• That any answer, or any department, used AI to write. Word counts show style; people used these words before ChatGPT and keep using them.
• That voice measurement, ScriptGrain's or anyone's, can tell whether a text was written by AI. Nothing in this study tests that.
• Why the writing changed. A new government and new ministers (July 2024), new staff, new drafting guidance, a changing mix of subjects (more questions about AI itself) and templates reused across answers could each move these numbers; this study measures the change, not its cause.
• Anything about oral questions, Hansard debates, committee evidence or the questions' content; or anything about who, a minister or an official, wrote a given word.

AI-associated words, year by year

Exploratory. Each year is the mean over departments with at least 5 answers that year; the pre-registered tests compare each department only with itself.

What changed, within the same departments

Pre-registered hypotheses and results

Written ministerial statements: the same six tests

Pre-registered as a secondary series, with the same rules as for answers. Statements are a census, not a sample, and there are fewer of them, so intervals are wider.

Department by department

Exploratory. Mean AI-associated words per 1,000 words in each period; a figure appears only where the department has at least 10 answers in the period.

By word list and habit

Within-department change in answers from 2019-22 to 2024-26, same departments as H1.

Commons and Lords separately

Checks on the weighting, the answers and the government

Exploratory robustness checks, each repeating the main comparison with one change. Conservative months: 1 January to 4 July 2024. Labour months: from 5 July 2024.

Written statements, and the questions themselves

Statements: the same six tests as for answers, secondary and pre-registered. Questions: the text MPs and peers wrote for the sampled answers, a comparison with no hypothesis attached.

Other bodies

Exploratory. The same comparison for answers from bodies outside the nineteen departments (1,875 answers), each body counting once.

Year by year

Exploratory; department-weighted means over the departments with at least 5 answers each year.

Limitations

• Written answers are drafted by civil servants, approved and given by a minister, and often reuse earlier answers, templates and shared lines; a numbers-only measure cannot separate the official's drafting from the minister's edit.
• Answers are a sample, not a census: a fixed-seed draw of up to 5,000 a year, weighted back to each department's output. Weighting corrects the design but not the exclusions that follow (short answers, duplicates), which fall unevenly across years.
• Many answers are very short, point to an earlier reply, or are mostly a list of figures. Texts under 50 words and mostly-numeric texts are left out, so the study describes substantive prose answers, not every answer.
• Word lists count spellings. These words rose in human writing for decades before ChatGPT, and topics shifted too: Parliament now asks about AI itself.
• Departments are joined into lineages by judgement. The business, energy, trade and science lineage mixes departments that did different work, and some work moved between departments in 2023. HMRC questions are answered through the Treasury and are counted there.
• Parliament was dissolved before the 2017, 2019 and 2024 general elections and prorogued between sessions (twice in 2019); no answers are given in those gaps, so months are not equally represented.
• The Conservative-months check covers about five months of answers (dissolution from 30 May 2024), so it is a weak test of whether the change of government explains the shift.
• Written questions and statements of the UK Parliament only, in English; oral questions and debates are not covered.

Competing interests

ScriptGrain sells writing-voice measurement and generation, and the analyst builds it. The hypotheses, eras, measures, sampling design and tests were committed to the public repository before the full measurement, every result is reported whether it passed or failed, and the data and scripts are public.

Reproducing this

• Pre-registration and scripts: docs/research/sgr-022-preregistration.md and scripts/experiments/sgr-022/ in the ScriptGrain repository (collect.mjs lists, samples and fetches the corpus, run.mjs measures and analyses it, write-entry.mjs builds this page from the summary).
• Every text is named by its Parliament question or statement ID in the data, so any answer can be re-read on questions-statements.parliament.uk and measured again.

Data

• Every text measured: kind, ID, house, department, date, era, sampling weight and every measure (CSV) (csv)
• Department by period: counts and mean of every measure, answers and statements (CSV) (csv)
• Year by year: department-weighted means for answers, statements and questions (CSV) (csv)
• Summary: sampling design, hypotheses, intervals, robustness checks, exclusions (JSON) (json)

Released under CC BY 4.0: free to reuse, including commercially, with credit to ScriptGrain and a link to this page.

Citations

• UK Parliament Written Questions and Statements API. Contains Parliamentary information licensed under the Open Parliament Licence v3.0
• ScriptGrain (2026). Did brand writing change after ChatGPT? (SGR-011)
• ScriptGrain (2026). Where did AI-isms come from? (SGR-007)
• Kobak, D. et al. (2025). Delving into LLM-assisted writing in biomedical publications through excess vocabulary. Science Advances

How to cite

ScriptGrain (2026). Do ministers' written answers sound different since ChatGPT? 42,082 parliamentary answers and 17,292 written statements, 2015 to 2026 (Study SGR-022, conducted 22 October 2026). Dataset licensed CC BY 4.0. https://scriptgrain.com/research/do-ministers-written-answers-sound-different-since-chatgpt

About the analyst

Jack Stovell has worked in finance and data for more than twelve years, building management reporting, forecasting and profitability models for advertising agencies and tech scale-ups, from SQL and Power BI reporting to board-level analysis. Since 2016 he has run Adapt Progress Evolve, an applied AI studio, where he builds and operates AI systems and data products: ScriptGrain's measurement of writing voice across 45 attributes, UK Spend, which brings 16.7 million rows of UK council spending into one queryable dataset, and more than thirty AI agents running in production. He designs ScriptGrain's studies and is accountable for every number in them.

• jstov.uk
• LinkedIn
• GitHub

More research

• Which classic authors write most like an AI? 1,129 authors, 6,293 books, measured
• Same Brief, Four Voices
• ChatGPT vs a measured voice