AI in Journalism 2026 Statistics (Updated September)

How many journalists and newsrooms use AI

1. 1

   82% of journalists use at least one AI tool, up from 77% a year earlier.

   Source: Muck Rack, 2026 ^[1]

2. 2

   47% of journalists use ChatGPT; Gemini use rose from 13% to 22% and Claude use doubled from 6% to 12%.

   Source: Muck Rack, 2026 ^[1]

3. 3

   56% of UK journalists use AI for their work at least once a week; 16% have never used it for a journalistic task.

   Source: Reuters Institute, 2025 ^[2]

4. 4

   49% of UK journalists use AI for transcription or captioning at least monthly, 33% for translation and 10% for first drafts of articles.

   Source: Reuters Institute, 2025 ^[2]

5. 5

   43% of UK business journalists use AI for work at least weekly, against 21% of lifestyle journalists.

   Source: Reuters Institute, 2025 ^[2]

6. 6

   42% of UK journalists under 30 use AI at least weekly, against 29% of those aged 50 and over.

   Source: Reuters Institute, 2025 ^[2]

7. 7

   Over 80% of journalists in the Global South and emerging economies surveyed by the Thomson Reuters Foundation use AI in their work.

   Source: Thomson Reuters Foundation, 2025 ^[3]

8. 8

   49% of newsroom respondents said in 2023 that their newsrooms were already using generative AI tools.

   Source: WAN-IFRA, 2023 ^[4]

9. 9

   97% of news executives rate back-end automation as an important use of AI for their organisation in 2026.

   Source: Reuters Institute, 2026 ^[5]

10. 10

    44% of news executives say their newsroom AI projects show promising results; 42% call the results limited.

    Source: Reuters Institute, 2026 ^[5]

11. 11

    75% of news executives expect AI agents to have a large or very large impact on the news industry (a forecast).

    Source: Reuters Institute, 2026 ^[5]

How many newsroom jobs have AI in the title

1. 12

   63 of the 127,602 journalists on file carry an AI word in their job title (0.049%), at 48 of 8,800 outlets.

   Source: JournalistLabs · count of September 15, 2026 · 127,602 editorial profiles at 8,800 media outlets

2. 13

   The BBC has the most AI job titles of any outlet on file, 5 of its 3,935 journalists, followed by Bloomberg with 3 of 1,785.

   Source: JournalistLabs · count of September 15, 2026 · 3,935 BBC and 1,785 Bloomberg editorial profiles

3. 14

   Of the 63 AI job titles, 27 describe running AI inside the newsroom, 21 describe covering AI as a subject and 15 are unclear.

   Source: JournalistLabs · count of September 15, 2026 · 63 AI job titles

4. 15

   34 of the 50 largest newsrooms on file have no one with AI in their job title.

   Source: JournalistLabs · count of September 15, 2026 · the 50 publishable outlets with the most editorial profiles

5. 16

   Magazines have 1.89 AI job titles per 1,000 journalists on file and broadcasters 0.27; newspapers have 0.47 and online outlets 1.07.

   Source: JournalistLabs · count of September 15, 2026 · 8 AI titles on 4,222 magazine profiles; 14 on 52,664 broadcaster profiles; 21 on 44,234 newspaper profiles; 18 on 16,885 online-outlet profiles · method

Do newsrooms have AI rules and training

1. 17

   Only 13% of journalists surveyed in the Global South and emerging economies said their employer had a policy on AI use.

   Source: Thomson Reuters Foundation, 2025 ^[3]

2. 18

   Almost 8 in 10 of those journalists had no established AI policy in their newsroom.

   Source: Thomson Reuters Foundation, 2025 ^[6]

3. 19

   Only 20% of newsrooms had guidelines on generative AI in 2023, and 49% let journalists use the tools as they saw fit.

   Source: WAN-IFRA, 2023 ^[4]

4. 20

   60% of UK journalists say their main outlet has AI guidelines on at least one issue; 32% say it offers AI training.

   Source: Reuters Institute, 2025 ^[2]

5. 21

   Only 14.1% of 504 media professionals surveyed in the Basque Country had received any AI training, mostly by teaching themselves.

   Source: Frontiers in Communication, 2025 ^[7]

What journalists think of AI

1. 22

   26% of journalists name unchecked AI as a top concern for the industry, up from 18% a year earlier.

   Source: Muck Rack, 2026 ^[1]

2. 23

   62% of UK journalists see AI as a large or very large threat to journalism; 15% see it as a large or very large opportunity.

   Source: Reuters Institute, 2025 ^[2]

3. 24

   60% of UK journalists are extremely concerned about AI's effect on public trust in journalism.

   Source: Reuters Institute, 2025 ^[2]

4. 25

   89.88% of 504 journalists surveyed in the Basque Country believe AI will considerably or significantly increase the risk of disinformation.

   Source: Journalism and Media, 2025 ^[8]

5. 26

   42% of journalists in the Global South survey were positive about the future use of AI.

   Source: Thomson Reuters Foundation, 2025 ^[3]

What audiences think of AI in the news

1. 27

   Only 12% of people across six countries are comfortable with news made entirely by AI, against 62% for news made entirely by humans.

   Source: Reuters Institute, 2025 ^[9]

2. 28

   Comfort with AI in news rises to 43% when a human leads the work with some AI help.

   Source: Reuters Institute, 2025 ^[9]

3. 29

   55% of people are comfortable with AI checking spelling and grammar in news; 19% are comfortable with artificial presenters or authors.

   Source: Reuters Institute, 2025 ^[9]

4. 30

   Only 33% of people think journalists always or often check AI outputs before publishing.

   Source: Reuters Institute, 2025 ^[9]

5. 31

   51% of people think news media use generative AI always or often.

   Source: Reuters Institute, 2025 ^[9]

6. 32

   Only 19% of people see AI labels on news daily, although 77% use news daily.

   Source: Reuters Institute, 2025 ^[9]

7. 33

   6% of people use generative AI to get news each week, up from 3% a year earlier.

   Source: Reuters Institute, 2025 ^[9]

8. 34

   54% of people saw an AI-generated answer to one of their searches in the past week.

   Source: Reuters Institute, 2025 ^[9]

9. 35

   27% of people across 48 markets want AI to summarise news for them and 24% want it to translate stories.

   Source: Reuters Institute, 2025 ^[10]

10. 36

    50% of U.S. adults expect AI to have a negative effect on the news people get over the next 20 years; 10% expect a positive effect.

    Source: Pew Research Center, 2025 ^[11]

11. 37

    59% of Americans expect AI to lead to fewer jobs for journalists over the next two decades; 5% expect more.

    Source: Pew Research Center, 2025 ^[11]

12. 38

    41% of U.S. adults say AI would do a worse job writing a news story than a journalist; 19% say it would do better.

    Source: Pew Research Center, 2025 ^[11]

13. 39

    97.8% of local news consumers surveyed want to know if a newsroom used AI, and 98.8% want a human involved.

    Source: Local Media Association and Local Media Foundation, 2025 ^[12]

14. 40

    47.6% of local news consumers surveyed are uncomfortable with journalists using AI, even when the output is checked, against 37.1% who are comfortable.

    Source: Local Media Association and Local Media Foundation, 2025 ^[12]

15. 41

    Only 26% of Americans polled for the National Association of Broadcasters trust information produced by AI; 68% say it is not trustworthy.

    Source: National Association of Broadcasters, 2025 ^[13]

16. 42

    76% of Americans polled for NAB are concerned about AI stealing or reproducing journalism and local news stories.

    Source: National Association of Broadcasters, 2025 ^[13]

17. 43

    Labelling headlines as AI-generated made 4,976 US and UK participants rate them as less accurate and share them less, whether the headlines were true or false and written by people or AI; the effect was about a third of the size of labelling them false.

    Source: PNAS Nexus, 2024 ^[14]

How accurate AI tools are with news

1. 44

   45% of AI assistant answers to news questions had at least one significant issue, and 31% had serious sourcing problems, in a study across 18 countries and 14 languages.

   Source: EBU and BBC, reported by NPR, 2025 ^[15]

2. 45

   Eight AI search tools gave incorrect answers to more than 60% of queries asking them to identify a news article; Perplexity got 37% wrong and Grok 3 94%.

   Source: Tow Center for Digital Journalism, Columbia Journalism Review, 2025 ^[16]

3. 46

   The 10 leading AI tools repeated false claims about news topics 35% of the time in August 2025, up from 18% in August 2024.

   Source: NewsGuard, 2025 ^[17]

4. 47

   The same AI tools declined to answer 0% of news prompts in August 2025, down from 31% a year earlier.

   Source: NewsGuard, 2025 ^[17]

5. 48

   A large language model correctly flagged 90% of false headlines, yet showing its fact checks to people did not improve their ability to tell true headlines from false ones.

   Source: Proceedings of the National Academy of Sciences, 2024 ^[18]

Frequently asked questions

How many journalists use AI?

Most do. 82% of journalists in Muck Rack's 2026 survey use at least one AI tool, and 47% use ChatGPT. In the UK, 56% of journalists use AI for work at least once a week, mostly for transcription and translation, according to the Reuters Institute.

Do newsrooms have AI policies?

Often not. Only 13% of journalists in the Thomson Reuters Foundation's Global South survey said their employer had an AI policy. In the UK, 60% say their outlet has guidelines on at least one issue, and 32% say it offers AI training.

Are people comfortable with AI-written news?

Few are. Across six countries, 12% are comfortable with news made entirely by AI and 62% with news made entirely by people, according to the Reuters Institute. Comfort rises to 43% when a human leads with some AI help.

How accurate are AI assistants about the news?

Not very. An EBU and BBC study across 18 countries found a significant issue in 45% of AI assistant answers to news questions. The Tow Center found eight AI search tools gave wrong answers to more than 60% of its queries, and NewsGuard found the 10 leading tools repeated false claims 35% of the time in August 2025.

How many newsroom jobs are dedicated to AI?

Few, going by job titles. 63 of the 127,602 journalists in the JournalistLabs count have an AI word in their title, at 48 outlets. The BBC has the most with 5, and 34 of the 50 largest newsrooms have none.

Sources

1. [1]Muck Rack. “State of Journalism 2026”. 2026.↑1 ↑2 ↑22
2. [2]Reuters Institute. “AI adoption by UK journalists and their newsrooms: surveying applications, approaches and attitudes”. 2025.↑3 ↑4 ↑5 ↑6 ↑20 ↑23 ↑24
3. [3]Thomson Reuters Foundation. “Journalism in the AI era: Opportunities and challenges in the Global South and emerging economies”. 2025.↑7 ↑17 ↑26
4. [4]WAN-IFRA. “New survey finds half of newsrooms use Generative AI tools”. 2023.↑8 ↑19
5. [5]Reuters Institute. “Journalism, Media and Technology Trends and Predictions 2026”. 2026.↑9 ↑10 ↑11
6. [6]Thomson Reuters Foundation. “Journalism in the AI era (report page)”. 2025.↑18
7. [7]Frontiers in Communication. “Artificial Intelligence Training in Media: Addressing Technical and Ethical Challenges for Journalists and Media Professionals”. 2025.↑21
8. [8]Journalism and Media. “Journalists' Perceptions of Artificial Intelligence and Disinformation Risks”. 2025.↑25
9. [9]Reuters Institute. “Generative AI and News Report 2025: How people think about AI's role in journalism and society”. 2025.↑27 ↑28 ↑29 ↑30 ↑31 ↑32 ↑33 ↑34
10. [10]Reuters Institute. “Digital News Report 2025”. 2025.↑35
11. [11]Pew Research Center. “Americans largely foresee AI having negative effects on news, journalists”. 2025.↑36 ↑37 ↑38
12. [12]Local Media Association and Local Media Foundation. “News consumers cautiously optimistic about AI use in news”. 2025.↑39 ↑40
13. [13]National Association of Broadcasters. “poll on AI and local news (conducted by OnMessage Inc.)”. 2025.↑41 ↑42
14. [14]PNAS Nexus. “People are skeptical of headlines labeled as AI-generated, even if true or human-made, because they assume full AI automation”. 2024.↑43
15. [15]EBU and BBC, reported by NPR. “News Integrity in AI Assistants”. 2025.↑44
16. [16]Tow Center for Digital Journalism, Columbia Journalism Review. “AI Search Has A Citation Problem”. 2025.↑45
17. [17]NewsGuard. “August 2025 AI False Claim Monitor”. 2025.↑46 ↑47
18. [18]Proceedings of the National Academy of Sciences. “Fact-checking information from large language models can decrease headline discernment”. 2024.↑48

Method

We searched Google for AI in journalism statistics, AI in newsrooms statistics and AI journalism statistics, read the 11 pages that ranked, and counted how many of them repeat each number. We traced each number to the organisation that produced it and checked it against that organisation's page on September 24, 2026. We searched OpenAlex, arXiv and Europe PMC for journal papers; few papers on newsroom AI report plain percentages in their abstracts, so papers make up a small share. Surveys of journalists differ in who they asked: the Thomson Reuters Foundation surveyed alumni of its own training programmes, Muck Rack surveyed journalists in its database, and the Reuters Institute's UK survey is broadly representative of UK journalists. The EBU and BBC study is read through NPR's summary, one of the participating broadcasters. The JournalistLabs numbers reuse our report on AI job titles and count job titles only, not profile headlines, fetched July to September 2026.

JournalistLabs numbers new on this page:

• Magazines have 1.89 AI job titles per 1,000 journalists on file and broadcasters 0.27; newspapers have 0.47 and online outlets 1.07. The report's per-outlet AI title counts summed by outlet type and divided by that type's editorial profiles. Wire agencies (2 titles on 5,448 profiles) are too few to rate.

All 48 statistics as a CSV: ai-in-journalism-statistics.csv.

Cite this page

JournalistLabs, “AI in Journalism 2026 Statistics (Updated September)”, updated September 2026, https://journalistlabs.com/research/ai-in-journalism-statistics/

To cite one number, link its anchor: each statistic has one, for example https://journalistlabs.com/research/ai-in-journalism-statistics/#stat-1.