# Grey literature: included items (source 4)

Search run 2026-09-26 per `LITERATURE-SEARCH-PROTOCOL.md`, source 4 only (practitioner surveys and industry reports, 2022–2026, English and German). Query and screening log: `grey-search-log.csv`.

**Already used by the report, noted but not re-extracted:** SessionLab State of Facilitation 2025 and 2026; Jigsaw "Facilitation in the AI Era" (2025); WEF Future of Jobs Report 2025; Doshi & Hauser 2024 (Science Advances); Leadership Strategies "2026 AI in Facilitation Survey" landing page.

**Leadership Strategies follow-up:** I found no separate public report with a method section. The landing page (`https://www.leadstrat.com/state-of-ai-in-facilitation/`) now states a sample of **256** "facilitators, L&D leaders, and executives", reports 0–5 frequency scores (in-session 0.75, prep 2.46, post-session 2.30), and gates the full report behind a lead form. It gives no fieldwork dates or recruitment method. The webinar page on Contrast returned HTTP 403. The course page (`/certified-ai-enabled-facilitator/`) repeats "74.2%", "48.4% flagged privacy and security" and "44.5% flagged quality and accuracy" without a method. These are marketing pages for a paid certification, which is a conflict of interest.

Items are ordered by relevance to the question: first facilitators, then meeting and event organisers, then AI's effect on meetings, then adjacent L&D populations. Where a number was taken from a page summary rather than a verbatim quote, the entry says so.

---

## 1. SessionLab — State of Facilitation 2024

- **Publisher:** SessionLab
- **Title:** State of Facilitation 2024 – Report and Expert Insights (2nd edition)
- **Date:** published early 2024; fieldwork September–November 2023
- **URL:** https://www.sessionlab.com/state-of-facilitation/2024-report/ (read in full as HTML)
- **Method:** online survey on Jotform, in English or Spanish. It was promoted to SessionLab newsletter subscribers and users, partners' networks, facilitation communities and social media. "The survey reached 3517 individuals and had a 27% completion rate resulting in 975 responses" (Appendix). Most questions were optional, so the base varies by question. Open answers were classified "with some help from ChatGPT 4.0".
- **n:** 975 respondents, 93 countries
- **Who was asked:** facilitators and workshop designers, the majority "from our own users and newsletter subscribers" (Appendix)
- **Key numbers:**
  1. "When we ran the survey in late 2023, 49.4% of facilitators were not yet using AI in their practice" (section *Focus on AI*; question: "What generative AI tools have you used in your facilitation practice?")
  2. "7% of facilitators used ChatGPT consistently during the design/planning phase of their workshops in 2023, with many more experimenting with it occasionally" (section *Frameworks & Tools → Tools*)
  3. An open question on how facilitators integrate AI got 257 responses. Idea generation was mentioned by 101, copywriting by 89, and "summarising large quantities of information … and transcribing and summarising conversations" by 53. "Very few responses indicate in-session use" (section *Focus on AI*, "Before a session / In session / After sessions")
- **Conflict of interest / caveats:** a vendor surveying mostly its own users, and it sells an AI agenda assistant. It acknowledges that its own tools are over-represented. **Internal inconsistency:** an expert-insight box in the same section says "65% of facilitators used AI tools in 2023", but 49.4% not using implies about 50.6% using. Use the 49.4% data figure, not the expert box. This is the 2023 baseline for the 2025 and 2026 editions the report already cites.

## 2. EITW for GCB / EVVC / DZT — Meeting- & EventBarometer Deutschland 2025/2026 (German)

- **Publisher:** Europäisches Institut für TagungsWirtschaft (EITW), commissioned by the German Convention Bureau (GCB), the Europäischer Verband der Veranstaltungs-Centren (EVVC) and the Deutsche Zentrale für Tourismus (DZT)
- **Title:** Meeting- & EventBarometer Deutschland 2025/2026 — Management-Summary
- **Date:** May 2026; organiser fieldwork mid-January to early March 2026
- **URL:** https://www.evvc.org/sites/default/files/2026-05/Meeting-%20und%20EventBarometer%202025%20-%20Ergebnispr%C3%A4sentation.pdf (PDF read in full)
- **Method (p. 63, *Projektdesign*):** online surveys of about 4,700 venues and of organisers. Organisers were reached through a selected mailing list of about 3,000 worldwide and a pre-screened panel of about 500 decision-makers for business travel and events in DE, US, UK, AT, CH, FR, NL and BE. Links were also spread via trade media, newsletters and social networks. The sample is partly self-selected.
- **n:** organisers ("Veranstalter") n=623 on the AI question; venues ("Anbieter") n=396
- **Who was asked:** organisers of events with at least 20 participants (companies, associations, agencies, public bodies)
- **Question (p. 51):** "Digitalisierung und KI revolutionieren auch die Veranstaltungswelt: Inwiefern nutzen Sie bereits künstliche Intelligenz in Ihrem Unternehmen?" (answer options: *Nutzen wir bereits / In Planung / Nutzen wir nicht / Weiß ich nicht*)
- **Key numbers (organisers, n=623, p. 51 "Veranstalter: Nutzung künstlicher Intelligenz"):**
  1. General AI applications ("Allgemeine KI-Anwendungen"): 72% already use, 18% planning, 8% do not use, 2% don't know
  2. Programme planning ("Programmplanung"): 46% already use, 29% planning, 22% do not use, 3% don't know
  3. Evaluation and reporting ("Auswertung & Reporting"): 46% already use, 32% planning, 19% do not use, 3% don't know
- **Conflict of interest / caveats:** no AI vendor involved; the commissioners are industry and tourism bodies. The respondents are event and conference organisers, not facilitators. "Programmplanung" is the closest item to agenda design but covers conference programmes too. This is the only German-language quantitative source found that reports AI use in planning group events.

## 3. Digital.ai — The 18th State of Agile Report

- **Publisher:** Digital.ai. Data collection and analysis by Regina Corso Consulting.
- **Title:** The 18th State of Agile Report
- **Date:** 2025 (fieldwork July 11 – August 20, 2025)
- **URL:** https://www.meriroos.ee/Stuff/Digital-ai-18th-State-of-Agile.pdf (third-party mirror of the PDF; read in full)
- **Method:** online survey; methodology note on the survey-questions page (p. 21)
- **n:** 349 agile practitioners "across diverse roles, industries, and regions"
- **Who was asked:** agile practitioners. Press coverage describes them as mainly agile coaches and consultants in large enterprises. The PDF itself only says "diverse roles".
- **Key numbers:**
  1. Q17 "In which areas is your organization using or piloting Agentic AI" (asked only of those reporting agentic AI experimentation or adoption; subgroup n not given). Under *Team enablement & facilitation*: "Facilitating retrospectives or sprint reviews" 12%, "Providing coaching or team feedback" 23%, "Supporting discovery/Customer research synthesis" 24% (p. 16). *Caveat:* values were read from a multi-column chart layout in extracted text, so check them against the rendered page.
  2. Q15 "Which types of AI tools are being used?" (AI users only): "General-purpose LLMs used manually by individuals" 65%, "Agile planning or prioritization assistants" 18% (p. 14)
  3. Q16 "Why are you using AI?": 77% "To save time or reduce manual effort". The narrative gives "summarizing retrospectives" as an example but does not quantify it (p. 15).
- **Conflict of interest:** Digital.ai sells agile planning and AI delivery software.

## 4. Microsoft — What Can Copilot's Earliest Users Teach Us About Generative AI at Work?

- **Publisher:** Microsoft (WorkLab / Work Trend Index special report)
- **Date:** November 15, 2023
- **URL:** https://www.microsoft.com/en-us/worklab/work-trend-index/copilots-earliest-users-teach-us-about-generative-ai-at-work
- **Method:** (a) survey of Microsoft 365 Copilot Early Access Program users, October 3 – November 2, 2023, n=297 (from a page summary, not verbatim); (b) controlled lab studies, including "The Missed Meeting" (study #2): 60 Microsoft employees watched a scripted 35-minute recorded Teams meeting "that included crosstalk, debate, and off-topic discussions" and wrote summaries, which were scored against a rubric of 15 meeting details (rubric detail from a page summary)
- **n:** survey 297; missed-meeting experiment 60
- **Who was asked:** early enterprise Copilot users (survey) and Microsoft employees (experiment)
- **Key numbers:**
  1. "Copilot users summarized the meeting nearly 4x faster (3.8), in 11 minutes and 13 seconds, compared to 42 minutes and 34 seconds." (Study #2, The Missed Meeting)
  2. "86% said Copilot makes it easier to catch up on what they missed."
  3. "84% said Copilot makes it easier to take action after a meeting."
- **Conflict of interest:** strong. The vendor is testing its own product, the experiment used its own employees, and the survey sample was early adopters. The experiment measures catching up on a meeting afterwards, not the conduct of the meeting.

## 5. Kolmogorov Law — AI notetakers in U.S. workers' meetings

- **Publisher:** Kolmogorov Law (commissioned); fieldwork on the Pollfish platform; syndicated by Stacker
- **Title:** "AI notetakers have sat in on 1 in 3 US workers' meetings, but only a third say they were asked first"
- **Date:** survey July 8, 2026; published July 16–17, 2026
- **URL read:** https://kvia.com/stacker-business-economy/2026/07/17/ai-notetakers-have-sat-in-on-1-in-3-us-workers-meetings-but-only-a-third-say-they-were-asked-first/ (the original at kolmogorovlaw.com returned HTTP 403)
- **Method:** online panel survey (Pollfish). Margin of error ±4.4% for the full sample and ±7.6% for the notetaker subsample.
- **n:** 500 employed U.S. adults aged 18+ (wage workers and self-employed); subsample of 167 who had an AI notetaker in their meetings
- **Key numbers:**
  1. "One in 3 employed Americans (33.4%) say an AI notetaker … has been present in their work meetings"; another 22.4% cannot say
  2. Of the 167 exposed workers, 34.7% "were always asked for permission" (36.5% sometimes, 25.1% never, 3.6% told only afterwards)
  3. 18.8% of all respondents discovered after the fact that a meeting had been recorded without their knowledge
- **Conflict of interest:** a law firm with an interest in consent and recording litigation. The questions frame consent. Only a syndicated copy could be read.

## 6. Microsoft — Will AI Fix Work? (2023 Work Trend Index)

- **Publisher:** Microsoft (Work Trend Index); fieldwork by Edelman Data x Intelligence
- **Date:** May 9, 2023; survey February 1 – March 14, 2023
- **URL:** https://www.microsoft.com/en-us/worklab/work-trend-index/will-ai-fix-work
- **Method:** online survey plus Microsoft 365 telemetry
- **n:** 31,000 full-time employed or self-employed workers in 31 markets (1,000 per market)
- **Key numbers (section "Top 5 Obstacles to Productivity" and "AI's Productivity Promise"):**
  1. "Most people say it's difficult to brainstorm in a virtual meeting (58%)"
  2. "next steps at the end of a meeting are unclear (55%)"; "hard to summarize what happens (56%)"
  3. "summarizing their meetings and action items (80%)". This comes from a list of tasks people would be comfortable delegating to AI. The full question stem was not captured verbatim.
- **Conflict of interest:** vendor research that sets up the case for its own meeting AI (Copilot in Teams). The data were collected before most respondents had access to meeting AI.

## 7. Zoom — Solving the collaboration paradox: survey shows how teams can save time with AI

- **Publisher:** Zoom; fieldwork by Morning Consult
- **Date:** fieldwork August 10–25, 2023
- **URL:** https://www.zoom.com/en/products/ai-assistant/resources/save-time-with-ai/
- **Method:** online survey
- **n:** 11,023 full-time knowledge workers (8,010 employees, 3,013 team leaders) in the US, UK, Ireland, Germany, France, Japan, Singapore and Australia
- **Key numbers:**
  1. "54% of employees say summaries and action items should often or always be sent after meetings, just 39% say they are actually sent"
  2. "73% of leaders and 55% of employees spend at least a few times a week sharing notes and action items with colleagues"
  3. "75% of leaders whose teams use AI say they collaborate better" (self-report, correlational)
- **Conflict of interest:** strong. The vendor sells Zoom AI Companion and the page is a product resource. Numbers were taken from a page summary; the quotation marks reflect the page wording as returned.

## 8. Global DMC Partners — Q3 2024 Meetings & Events Pulse Survey

- **Publisher:** Global DMC Partners (network of destination management companies)
- **Date:** press release September 16, 2024; fieldwork June 27 – August 9, 2024
- **URL:** https://s24.q4cdn.com/538403808/files/doc_news/Global-DMC-Partners-Releases-Q3-2024-Meetings--Events-Survey-Results-2024.pdf (press release PDF; full results not opened)
- **Method:** pulse survey. The administration method is not stated.
- **n:** 165 respondents, "primarily event professionals from the U.S., Canada, the U.K., and Europe … corporate planners, agency planners, and freelancers"
- **Key number:** "AI adoption in the industry has grown, with 48 percent of respondents now using AI in their work, up from 30 percent at the end of 2023. Chatbots like ChatGPT and CoPilot are among the most used tools." (section "AI")
- **Conflict of interest:** low (no AI product). The population is event planners, not facilitators. The sample is small and the full questionnaire was not seen.

## 9. ATD — AI in Instructional Design: Transforming Workflows and Content Creation

- **Publisher:** Association for Talent Development, sponsored by Clarity Consultants
- **Date:** August 4, 2025 (press release)
- **URL read:** https://www.prnewswire.com/news-releases/atd-research-ai-tools-are-benefitting-instructional-designers-302521023.html (the full report is a paid ATD product and was not opened)
- **Method:** survey. Fieldwork dates are not stated in the press release.
- **n:** 232 talent development professionals
- **Key numbers (press release; from a page summary, not verbatim):**
  1. Nearly two-thirds of instructional designers began using AI tools within the last year
  2. 37% say AI tools "greatly reduced the amount of time it takes to design a course"; 70% say AI "improves the quality of their course design"
  3. 20% of instructional designers do not use AI, mainly citing "lack of trust in AI-generated content"
- **Conflict of interest:** sponsor is a staffing and consulting firm; no AI product. The population is adjacent: designers of training, not facilitators of live group sessions.

## 10. CIPD — Learning at work 2023: Survey report

- **Publisher:** CIPD (fieldwork with YouGov acknowledged)
- **Date:** June 2023; fieldwork January–February 2023
- **URL:** https://www.cipd.org/globalassets/media/knowledge/knowledge-hub/reports/2023-pdfs/2023-learning-at-work-survey-report-8378.pdf (PDF read in full)
- **Method:** online survey; 91% of respondents UK-based
- **n:** 1,108 complete responses "from individuals who have a responsibility for supporting learning at work"
- **Key numbers:**
  1. Figure 12, "Digital technologies used to support L&D now and planned use for the coming year (% of respondents)": "Artificial intelligence tools (eg ChatGPT)" 5% currently in use, 6% planning to use in the coming year; base n=1,108 (Section 4 "What methods and media are in our L&D kitbag?", pp. 18–19)
  2. Voluntary-sector sub-sample: "two per cent are working with generative AI (in February 2023) and 3% planned to start" (voluntary sector summary, near the end of the report)
- **Conflict of interest:** none (professional body). The data are an early-2023 baseline taken weeks after ChatGPT's launch. The population is L&D functions, not facilitators.

---

## Counts

| | n |
|---|---|
| Queries run (EN 27, DE 13) | 40 |
| Items screened | 50 |
| Already in report (noted, not re-extracted) | 3 (SessionLab 2025, SessionLab 2026, Leadership Strategies landing page; Jigsaw, WEF and Doshi & Hauser did not surface as separate hits) |
| Excluded | 37 |
| Included | 10 |

**Excluded by reason**

| Reason | n |
|---|---|
| Off-topic: no data on facilitators, meeting organisers or group sessions (general workforce, HR, students, learners, non-AI) | 18 |
| Could not be opened, paywalled, gated, or primary source not accessed | 10 |
| Vendor marketing or claim without method or sample | 5 |
| Not grey literature (academic paper; passed to sources 1–3) | 2 |
| One-to-one setting (coaching), excluded by protocol | 1 |
| Published before 2022 | 1 |

**Gaps worth stating in the report:** no facilitation association (IAF, INIFAC, IAP2) has published a member survey on AI. No German survey of Moderatoren or trainers with a stated method and n was accessible: managerSeminare's BranchenBlick polls publish no n, and the TH Augsburg/changeXperten study is qualitative and paywalled. Every direct survey of facilitators on AI found comes from a vendor or a training firm with a product to sell (SessionLab, Leadership Strategies, Digital.ai).