Grants:Programs/Conference Fund/WikiCredCon 2026:"Addressing Reliability in an Era of Information Erosion"/Final Report

Hacks/Hackers, Inc. User:NevinThompson

WikiCredCon 2026: "Addressing Reliability in an Era of Information Erosion"

06 November 2025 - 30 June 2026

Report ID: 14020

Report status: Accepted

Report due date: 30 July 2026

Grant ID: G-CF-2508-19866

Amount funded: 81300 USD, 81300 USD

Amount spent: 86640.57 USD

Conference Fund Final Report

Part 1: Understanding your work

1. What do you think will be the short and long-term impact of this conference on participants, Wikimedia contributions, and your broader community? (required)

Short-term: participants learned about new and updated tools and methods they can incorporate into their own work. Examples include Cite Unseen’s citation-reliability icons, ideas for using structured AI-assistant for expanding thin biography leads, and ClanDi’s editor-training model, whose first pilot cohort launched within weeks of the conference. See Q3 for a list of tools and methods participants discussed.
Unexpected short-term outcome: WikiCredCon provided a space for media to engage directly with and then report on the community and the Foundation. The New York Times published "Wikipedia Is Battling for the Soul of the Internet" (July 5, 2026), less than three months after the conference, covering many of the same themes discussed at WikiCredCon, and quoting Bernadette Meehan, who had given a fireside chat at the event. WikiCredCon, which helped coordinate this, has a subtle shoutout in the article (as the Verge September 2025 article which started with reporting at WikiCredCon 2025).
Short/medium-term: WikiCredCon raises the profile of and energy around credibility issues in the Wikimedia ecosystem, including through the presence of leading community thinkers. For instance, Britta Gustafson (User: Dreamyshade), who was recently named as a 2026 Wikimedia Laureate, attended and presented at WikiCredCon to invite feedback on efforts around “deceptive” websites. This is one example among many that the conference will continue to attract leaders from the Wikimedia community to address credibility and information integrity.
Long-term: WikiCredCon connects different communities across the platform to address reliability online. For example, WikiLatinos’ ClanDi project (since renamed as WikiLatinos Learning Platform, or WLP), as discussed by Oscar Costero, is training Latino reproductive-rights activists as Wikipedia editors, and the Bay Area Wikipedians User Group is a confirmed recruitment pipeline: 3 of the 22 post-conference survey respondents (out of 95 attendees) named it directly as their entry point.
Long-term: WikiCredCon contributed to the continued relevance of Wikimedia beyond the movement itself. Joan Donovan, a prominent researcher and commentator on media manipulation and platform accountability, delivered the Sunday keynote, "Addressing Reliability in an Era of Information Erosion," situating Wikipedia's credibility challenges within the broader disinformation research landscape.
Long-term: the conference format is repeatable, not something that needs redesigning each year. WikiCredCon 2026 was the second annual event at the Internet Archive, and refined the initial 2025 format. The lightning-talk-into-unconference format has now worked across two years and can be reproduced easily by experienced event producers.
Long-term: the citation-tooling ecosystem is converging. Cite Unseen, the Internet Archive Reference Explorer, the Archive Citation Engine, and WikiSignals were all presented as working, deployed tools. Liam Wyatt’s shared-citations proposal named Cite Unseen specifically as a candidate for shared infrastructure.

Notably, Wyatt’s proposal at WikiCredCon sparked exploration of a smaller, related proof-of-concept currently called the “ReCite” system. ReCite — a reliable, reusable database for Wikipedia citations — is currently being developed by Connie Moon Sehat with WikiCred support. Since WikiCredCon in April 2026, productive exchanges around a front-end solution using Suggestion Mode have occurred with WMF Product Team and WMDE Technical Team members. The possibilities for backend have been and continue to be explored externally with Internet Archive and Zotero, and internally with members of the WikiData/WikiBase Technical Team.
Finally, improvements to the reorganization of the English Wikipedia Reliable Sources/Perennial Sources (RSP) list ([1]) continue to be nurtured through WikiCred community efforts.

These examples demonstrate how WikiCredCon and WikiCred serve a key coordinating role around the quality of on-wiki content through off-wiki organizing.

2. Did this conference improve participants’ ability to learn new skills, knowledge or build relationships in their Wikimedia work? (required) Yes

3. If yes or no, please briefly explain why, and what evidence or feedback informed your reflection. (required)

Survey responses confirm skill and knowledge transfer: attendees left with specific projects to continue, including LLM-aided citation verification, shared citations/WikiCite work, source-claim checking, and engaging new contributors. Attendee feedback was positive.

As one survey respondent noted, “Amazing time, well organized. Expanded all our abilities to improve credibility resources in our individual tools and work."

WikiCredCon conference improved participants' ability to learn new skills and knowledge through the following talks and workshops:

Cite Unseen (SuperHamster & SuperGrey): participants learned how to use and contribute to this globally deployed Wikipedia tool, which adds reliability icons to citations across multiple language Wikipedias. At least 600 editors have installed it; over 40 have used the crowdsourced categorization feature.
Internet Archive Citation Engine (Daniel Erasmus, IA Europe): participants were shown a live beta tool that extracts claims from Wikipedia paragraphs and matches them against 40 million academic articles, flagging supported or refuted claims.
Cite Unseen Filtering Dashboard: editors learned how to use the new filtering UI to identify and prioritize unreliable sources within specific Wikipedia articles.
Case study: Experimenting with AI-expanded biographies (Rosie Stephenson-Goodknight): participants explored a structured prompt method for experimenting with generative AI to expand short lead paragraphs on underrepresented women's biographies, where ORES scores improved in Start-class and C-class articles.
Gender bias detection tools (Netha Hussain): participants were shown two live tools, a bias pattern detector (48 patterns), and a Wikidata red-list explorer, for finding missing women's biographies.
Deceptive website identification (Britta Gustafson): editors learned to use Spam Check on Toolforge to identify and trace fake reliable-looking websites.
Productive friction framework (Bobbi Rakova): participants were introduced to a sociotechnical design concept for integrating AI tools into Wikipedia's editorial workflow.
Shared citations proposal (Liam Wyatt): participants gained a concrete mental model for how a centralized WikiCite database could work.
Resource building workshop ("How do we address today's issues around manipulated content and reliable sources?"): Guided by Connie Moon Sehat, attendees examined how to check for manipulated content and reliable source quality on Wikimedia, as government data is deleted or becomes degraded, and as generative AI creates more content online. Workshop attendees came up with practical methods to address this challenge and issues that will need addressing. Detailed notes about their conversations were published publicly following the workshop. (see 4. Documentation).

Survey respondents said they were able to form new relationships at WikiCredCon:

One newcomer respondent described feeling welcome and connecting with others working on similar problems despite knowing no one beforehand.
Unstructured time was a key part of WikiCredCon’s success: Networking opportunities scored 4.5 out of 5 across 22 survey responses, and seven of the 22 respondents (32%) named networking, connecting with people, or hallway time explicitly under what went well; two respondents specifically asked for more unstructured time next year.
Existing relationships also drove new attendance: one participant said colleagues who had attended in prior years gave her "FOMO" about coming this year.
Three respondents named the Bay Area Wikipedians User Group as the channel through which they first heard about the conference, indicates BAWUG functions as a local pipeline to WikiCredCon.

4. Documentation of your impact (optional). Please use the two spaces below to share files and links that help tell your story and impact. This can be documentation that shows your results through testimonies, survey results, videos, sound files, images (photos and infographics, etc) social media posts, dashboards, etc.

WikiCredCon 2026 schedule: [2]
Lightning talks: [3]
Notes - Resource Building Workshop: [4]
Short WikiCred.org event recap: [5]
Post-conference survey results: WikiCredCon 2026 Survey + Survey Responses.pdf
Video recordings or lightning talks: [6]
Press coverage: [7]

Wikipedia Is Battling for the Soul of the Internet - The New York Times.pdf

5. To what extent do you agree with the following statements regarding the work carried out with the support of this Fund? (required)

Our efforts during the Fund period have helped to...
A. Conduct the event in a safe and friendly environment.	Agree
B. Diversity of participants contributing to Wikimedia projects, including geographic, gender, and other relevant demographics.	Agree
C. Newcomers to the movement gained useful knowledge, skills, or relationships that helped them start or continue contributing to Wikimedia projects.	Strongly agree
D. Participants experienced a more inclusive and connected community, and felt better able to collaborate with others after the event.	Agree
E. Participants found the event content relevant and useful for their Wikimedia work.	Strongly agree
F. The event contributed to the creation or improvement of Wikimedia content from underrepresented perspectives.	Strongly agree
G. The event contributed to the improvement of the Wikimedia technical ecosystem through tools, demos, hacking etc.	Strongly agree
H. Participants have a plan for continued engagement for contributing to Wikimedia projects after the event.	Strongly agree
I. Created new partnerships or relationships with other non-Wikimedia groups as supporters and stakeholders.	Strongly agree
J. Organizing and executing the conference was a positive experience for the COT and sub-committee members.	Strongly agree

6. Documentation of resources (optional). Please share resources that would be useful to share with other Wikimedian organizations so that they can learn from, adapt or build upon your work. (please include links to any blogs, commons files here)

We are happy to provide resources for future organizers — please contact us.

7. As an organizer, what would you like to share as key learnings which could help future organizers, what information or learnings would you like to pass on? (required)

Catering is community infrastructure. Good food keeps energy high, encourages people to linger and talk, and signals that organizers respect attendees' time and wellbeing. Source locally, accommodate dietary needs proactively, and budget for it. It is well worth the investment if you can buy from local vendors.
Protect unstructured “hallway time.” Survey respondents consistently named unstructured conversation as the most valuable part of the conference. Design the schedule with more breaks so hallway time is protected and plentiful.
Build the unconference into the whole weekend. Putting breakout topic ideas on the wall from the very start of the conference during dinner Friday, where conversation threads begin and ideas germinate – rather than on Saturday evening, as scheduled — would help themes evolve organically over two days. As well, all participants would better understand the trajectory and goal of the weekend: an unconference on the Sunday. The unconference playbook should also be ready and internalized by organizers, so execution happens quickly on the final day. Start with 15–20 minutes of structured brainstorming, then transition to action. Designate a motivated facilitator for each group in advance, with a clear charge. Have a plan for capturing unconference topics and sharing with attendees later.
Clearly explain code of conduct guidelines leading up to the conference, and then reinforce expectations at the start of each day. Since WikiCredCon is in part aimed at growing the community, not every attendee will know unspoken Wikimedia cultural practices. These include what different-colored lanyard means, but also other conventions.

For example, it is important to explain and tactfully enforce community rules around transparency, notably identifying oneself (name or username), and one’s affiliation. Thought should be put into how to deescalate and resolve code of conduct violations.

Plan for changing budget inputs. In particular, accommodation costs can inflate from year to year (or even from the start of the funding agreement), and from when a proposal is submitted to when lodging is booked. Perceptions of personal safety can also influence how many participants are prepared to cross borders or travel through airports.

Part 2: Your Impact and Learnings

8. What were the major challenges or limitations, if any, that impacted attendees or the organizers? What would you do differently next time? (required)

We addressed the two main learnings from 2025 (too little unstructured time; need for more careful Q&A formats): the 2026 schedule had more breathing room, and Q&A was moved to a separate room, which respondents praised.
Fitting in 20+ lightning talks made it challenging to preserve valuable unstructured time. Consider a pre-event online space to share research interests, and structured mixer activity on the Friday evening (e.g., instead of lightning talks, a speed-dating activity with elevator pitches for their research). Our solution was to schedule lightning talks on the Friday evening, and during mealtimes; this did subtract from unstructured time.
Saturday was a full day, but was still considered by attendees to be high-value, and an improvement over 2025. Despite improving the schedule's breathing room versus 2025, several survey respondents noted that Saturday was a long day. For 2027, consider increasing unstructured “hallway time,” ending Saturday earlier (or using early Saturday evening as an unstructured activity) and building in a longer mid-afternoon break. However, these may be minor tweaks next time. As one survey respondent said: “Seriously, do not change the schedule. It was sustainably paced with quick lightning talks, higher context workshops/panels, and room for unstructured connections."
The unconference needed more structure. Suggestions: put unconference topic ideas on the wall throughout the weekend, not just at 8:30pm on Saturday; designate a facilitator for each breakout in advance.
Shared accommodation costs came in higher than in prior years. Airfare spend was under budget, due to lower-than-expected travel-subsidy demand, driven partly by strong local Bay Area turnout (such a strong group!) and fewer confirmed returning out-of-town attendees than in previous years, in turn due in part to concerns about immigration enforcement at U.S. airports. Communicating closely with WMF team, we shifted funds from airfare to lodging and conference materials. This reallocation matched actual attendee demand and did not affect the number of scholarship attendees.
Event security is now a key component of event planning, and must be resourced adequately: Logistics, communication, programming, travel and lodging, travel scholarships and other activities all generally receive adequate attention when designing and executing a proposal. However, event security must be recognized as a core activity, and should be resourced adequately. This includes budget for sourcing, evaluating, selecting, a vendor, but also for developing and implementing a security plan, and managing contracted security staff. Security protocols should be established with the funder at the very beginning of the grant award period, with the funder playing the guiding role to help organizers develop a plan.

9. Can you describe your learning on the scholarship selection? Please include any information on how the team was able to bring the right participant group, particularly for underrepresented communities or newcomers. (required)

Scholarships were awarded according to several criteria:

Returnees from 2025 (all had contributed in some way to the 2025 program) were given priority.
Wikipedians and Wikimedia community members from across North America were invited to obtain a scholarship.
BAWUG members from outside of San Francisco proper and easy commute were offered a scholarship.
Some invited presenters received support for travel, or lodging.

Due to the growing prominence and awareness of WikiCred and WikiCredCon, in future years organizers should implement a scholarship application and application review process.

10. Please describe your learnings regarding the conference program, including how sessions and activities were designed and selected to benefit attendees. (required)

The program's five theme areas — improving media literacy, enhancing citation transparency, tackling disinformation, empowering editors, and long-term collaboration with outside disinformation experts on Wikipedia's guidelines — were set by the organizing team ahead of the call for proposals and published on the conference's Meta-Wiki page. These five conference themes were based on what the pre-conference Community Engagement Survey, conducted in September 2025, had already surfaced as priorities.
The single-track, lightning-talk-into-unconference format is a continuation of the format used at WikiCredCon 2025 and, further back, the original WikiCredCon convened at MIT in 2019 alongside WikiConference North America. In the previous 2025 post-event survey results, attendees had said they preferred collaborative, hands-on activities and unstructured time over conventional presentations. In response, we replicated that format in 2026 with refinements rather than redesigned from scratch.
Speakers and sessions were then selected through an open call for proposals, which was launched in February 2026, about three months before the start of the conference. Promotion occurred at WikiCred meetings and in the WikiCred newsletter, on relevant Wikimedia community newsletters, and by direct outreach to possible presenters. Around 30 speakers submitted proposals to the organizing team, several offering two to four possible topics each, and the organizing team evaluated each before creating the final program for the weekend.
A key part of the program was both the quality of the talk proposals, and the desire by most attendees to propose and deliver a talk. As a result, we created time on the Friday evening for lightning talks. However, we ended up scheduling more talks during mealtimes, and also on Sunday. By discussing ideas, methods and tools, and by helping presenters broadcast their interests to the room, lightning talks added to the quality of the event, although it was challenging to fit in hallway time.
Program selection specifically favored sessions that served newcomers and underrepresented communities. Oscar Costero’s talk on ClanDi, which trains Latino reproductive-rights activists as Wikipedia editors, and Margarita Trovato’s session on Wikimedia Argentina’s Argentinian perspective on information integrity were both selected in part because they broadened the program past a US-only, English-Wikipedia framing. Emeline Brule’s talk on editor inclusivity, drawing on her experience mentoring women editors, and Netha Hussain’s gender bias detection tools were selected as directly relevant to the diversity goals in the funding proposal, not as separate additions to a technical program.

11. Can you share your observations about how the program impacted attendees’ knowledge, skills, collaboration, or future engagement within the Wikimedia movement. (required)

LiAnna Davis's Wiki Education talk highlighted a concrete way to address AI-generated content on the platform. Her data showed Pangram-based detection and training cut generative-AI-drafted content reaching live articles from a projected 25% down to 5% in a fall 2025 study, prompting discussion of applying that approach more broadly. Multiple survey respondents named it among the conference's most valuable sessions.
Oscar Costero presented a WikiLatinos project that gave attendees a transferable recruitment model. Train people who already hold subject-matter trust within a community to become editors, rather than recruiting generalist volunteers. The project, ClanDi (now WLP), applies this to Latina reproductive-rights activists to close the English/Spanish gap in reproductive-health coverage, with a first pilot cohort already launched.
Netha Hussain discussed two tools they can use immediately to address gender bias on Wikipedia. The first, a browser-based bias detector, flags patterns like patronizing tone, diminutive language, unnecessary gendering, and appearance-focused framing using rule-based matching (48 patterns, no AI), and suggests rewrites with a direct link to fix the article. The second, the “Women in Red Wikidata explorer tool,” empowers editors to search by occupation and country to find notable women with Wikidata entries but no English Wikipedia article yet, helping target edit-a-thons toward the biggest gaps.
Jamie Flood provided insights on teaching students information literacy and source credibility. Jamie’s session covered web archiving tools, trusted data sources (federal, state, and local), and Wikipedia-based class projects like updating outdated citations or writing articles. Flood also discussed how AI is complicating research on campus, with both students and faculty bringing in hallucinated references, and shared her openly licensed Wikipedia Guide & Curriculum for more assignments and resources.
Rosie Stephenson-Goodknight presented a potentially replicable, no-cost method for improving biography quality at scale. She discussed the results of a study, where a structured AI prompt expanded short leads on 75 of her own articles, improving the ORES quality rating in 73 of 75 cases, at under 7 minutes of volunteer time per article.
Phibeatrice reframed editor diversity itself as a credibility issue, separate from tool-building. Editors closer to a culture, language, or tradition catch knowledge gaps and inaccuracies others miss. The point landed personally, too: despite years as an editor, this conference was the first time they had met a fellow Vietnamese Wikipedia editor in person.
Elan Hourticolon-Retzler (Sources Please) gave attendees a reframing for tackling misinformation at scale: treat the underlying claim, not the individual post, as the unit worth verifying. He cited a striking asymmetry: professional fact-checkers spend 15–45 minutes debunking a claim that AI tools can regenerate in about three seconds, meaning a thousand-fold effort gap that post-by-post moderation tools can't close. Elan noted that Wikipedia's own sourcing model, in place since 2005, already solves this by requiring editors to cite verifiable sources rather than assert truth directly, which is in fact a template other platforms fighting misinformation could learn from.
Rebecca MacKinnon, Sally Lehrman, and Luis Villa gave attendees a shared framework for locating WikiCred's work within larger political and technological debates. MacKinnon on credibility and human rights, Lehrman on news trust amid AI and polarization, and Villa on the role of LLMs in free knowledge.
Margarita Trovato provided Wikimedia’s Argentina perspective on operationalizing information integrity, drawing on Wikimedia Argentina's experience.

12. How do you plan to build on the key successes of this conference? (required)

Continue monthly WikiCred online meetups. The conference generates energy that needs a regular cadence of follow-up to sustain it.
Facilitate a WikiCred track at WCNA 2026. This was a metric target in the grant (3 track proposals by WikiCred participants).
Pursue WikiCred affiliate status. The community now has two years of annual events, a regular meeting cadence, a clear thematic focus, and growing membership.
Advance the shared citations / WikiCite infrastructure conversation. The convergence of tools at WikiCredCon 2026 (Cite Unseen, Reference Explorer, IA Citation Engine, WikiSignals) suggests the community is ready to push this forward with WMF.
A constraint on follow-up: organizers must continue to build capacity in WikiCred to continue WikiCredCon.

13. Did you collaborate with any WMF teams to support the event? If yes, please indicate which teams you worked with and the type of support they provided. (required)

WikiCredCon received assistance from the following people and teams:

Andrea Romanos Roca, Movement Communications Specialist, Wikimedia Foundation (project management)
Elena Lappen, Movement Communications Manager, Wikimedia Foundation (event support; liaison with WMF attendees)
Barbara Levinsky, Communications Specialist, WMF (communications support)
Rachit Sharma, Manager (Global Events & South Asia), Wikimedia Foundation (project management)
A total of 13 WMF representatives attended WikiCredCon to spend three days with conference attendees, including the new CEO, Bernadette Meehan.
WMF staff also participated in two Chatham House Rules round tables, answering detailed questions about Foundation policy and strategy around AI and other topics, as well as plans for the future in the context of information integrity and credibility.

Part 3: Metrics

14. Open Metrics reporting. In your application, you defined some open metrics and targets (goals). You will see a table like the one below with your metric in the title and the target you set in your proposal automatically filled in.

Open Metrics Summary
Open Metrics	Description	Target	Results	Comments	Methodology
Community focus (1/2)	Total attendees	50	95	Nearly double the target of 50: 86 ticketed registrants plus roughly 9 walk-in attendees (press, additional Internet Archive staff) not captured by ticketing but vetted through security as trusted attendees. Day-of check-in (badge scans) recorded 80, likely an undercount of attendees and organizers who didn’t scan in.	Pretix; Final registration records, including paper sign-in sheet.
Community focus (2/2)	Total Event Satisfaction Score regarding Importance and Focus of the Convening (as a percentage)	85	92	Based on 22 survey responses across 5 questions (content, relevance, logistics, networking, overall). Overall experience alone: 92.7%. 21 of 22 respondents (95%) rated overall experience 4 or 5 out of 5.	Post-conference online survey (22 responses, conducted week following conference); see "Uploaded Documents and Files" for full anonymized responses.
Next steps	Average monthly WikiCred call attendees (over 6 months)	6	8	Average 8 per subsequent monthly online meeting (circa July 29, 2026)	WikiCred meeting rolling notes doc.
Deeper collaboration	WikiCred track proposal suggestions at WCNA 2026	3	3	As of July 29, WCNA 2026 program has not yet been announced. The result of 3 reflects informal WikiCred-member reports of planned submissions.	Discussions at online WikiCred meetings.
N/A	N/A	N/A	N/A	N/A	N/A

15. Based on the metrics you reported in your proposal and the eventual results, what do these results suggest about the impact of this conference for your community and Wikimedia projects? (required).

Growing attendance suggests WikiCredCon is pulling in an expanding base of people working on Wikipedia credibility, not just repeating the same room each year. The Open Metrics target was 50; actual attendance reached 95, 90% over target, and 36% above the 70 who attended WikiCredCon 2025. That growth can be attributed to the quality of the 2025 conference (strong support from the Foundation and the Internet Archive, a talk by Craig Newmark, and the energy of the WikiCred community itself), a sustained awareness campaign ahead of 2026, and BAWUG's outreach to Bay Area Wikipedians in 2026. This indicates a widening, not just larger, community engaging with credibility work year over year.
Satisfaction rising alongside that growth suggests the program can scale without diluting the experience that turns attendees into contributors to the platform. Satisfaction landed at 91.5% against an 85% target, even as attendance increased over 2025. (The 91.5% figure averages five broader survey questions.) This attendee satisfaction can be attributed to lessons learned from 2025, plus increased interest and support from the Foundation, the Internet Archive, and BAWUG.
Early data on WikiCred attendees and WCNA 2026 proposals suggests the conference's energy is carrying into sustained, year-round activity. Monthly WikiCred call attendance has run 8-15 people so far, above the target of 6 (final count still being tracked through October 2026). WCNA 2026 is projected to include at least six WikiCred-related talks, consistent with the high volume of talks and lightning talks at WikiCredCon 2026 itself.

16. Please indicate if you applied any of the following survey and registration tools. Please select all that apply. (required)

2. Post-event participant survey,
5. I used other survey or registration tools

17. Please share the result(s) with us, provide the link(s) or summarize the main result(s) and insight(s) from them. (optional)

22 survey responses collected. Average scores by question (out of 5): session content 4.5, topic relevance 4.4, organization and logistics 4.8, networking opportunities 4.5, overall experience 4.6.
Combined across all five questions, the satisfaction score is 91.5%; on the overall experience question alone it is 92.7%. 21 of 22 respondents (95%) rated overall experience 4 or 5 out of 5.
Top strengths named in open-ended responses: lightning talks, the Internet Archive venue, food, hallway and networking time, and this year’s schedule pacing.
Top improvement areas: Saturday’s length, cited by 5 of the 16 respondents who answered the improvement-areas question; the unconference’s lack of structure; and a handful of sessions that ran long relative to their content.

18. If you used other forms, please share them with us, as these forms might be useful for others to use. (optional)

Post-conference survey results: WikiCredCon 2026 Survey + Survey Responses.pdf

19. Based on the findings of the post event survey what were the main findings and take aways, how can they be helpful for future organisers. (required) Several from the survey are directly actionable for future organizers.

First, some attendees considered Saturday to be a somewhat long day: 5 of the 16 respondents who answered the improvement-areas question raised it specifically. Generally, there was great demand to include lightning talks, which claimed some of Saturday’s free time.

Survey feedback here was consistent with verbal feedback gathered from attendees and organizers during the event itself, reinforcing that Saturday’s length is worth addressing structurally in 2027 rather than through incremental schedule trims.

Second, the unconference and breakout sessions need a facilitation plan set before the event. Several respondents suggested designating a facilitator per breakout group in advance and posting topic ideas on the wall earlier in the weekend rather than only Saturday night, consistent with the recommendation already in Q7.

Some respondents noted overlap between similar tool presentations and suggested grouping talks by shared challenge rather than presenting each tool on its own; this may help open up more unstructured “hallway time” while still including as many relevant talks as possible.

Finally, conference attendees find WikiCredCon very valuable. The event, which one survey respondent described as welcoming, provides a space for participants to meet and share ideas about improving media literacy, enhancing citation transparency, and tackling disinformation, while networking and building connections for future collaborations. WikiCredCon is valuable because Wikimedians surface issues, ideas, and concepts that can be discussed with the Internet Archive, Wikimedia Foundation staff, grassroots members of the Wikipedia community, and researchers and technologists, plus newcomers to the conversation and the community.

Part 4: Financial reporting and compliance

20. & 21. State the total amount spent in your local currency. (required)

86640.57 USD

22. State the total amount spent in USD.

86640.57 USD

23. Report the funds received and spending in the currency of your fund (required). You can upload it or give a link to it below.

FINANCIAL REPORT WikiCredCon 2026 Budget.xlsx

We have uploaded an xlsx expense document to Fluxx, above, as part of this report. Let us know if you need a link as well, and we will add here immediately.

As required in the fund agreement, please report any deviations from your fund proposal here. Note that, among other things, any changes must be consistent with our WMF mission, must be for charitable purposes as defined in the grant agreement, and must otherwise comply with the grant agreement.

24. If you have not already done so in your budget report, please provide information on changes in the budget in relation to your original proposal. (optional)

The amount awarded ($81,300) exceeded the amount requested ($75,000). This is because additional budget was added for new security vendor requirements. Final actual spending came to $86,640.57, an overage of $4,908.57 that Hacks/Hackers covered from its own program funds beyond the grant.
Accommodation ran $6,638.61 over budget ($23,588.61 actual vs. $16,950 budgeted). Seeing spare capacity in the travel line, the team shifted from farther, cheaper Airbnbs to housing closer to the venue, which cost more per night.
Travel came in $7,194.72 under budget ($11,505.28 actual vs. $18,700 budgeted). Several scholarship recipients turned out to already be based in the Bay Area, so fewer flights and less flight assistance were needed than planned.
Staff costs ran $6,459.23 over budget ($26,084.23 actual vs. $19,625 budgeted), reflecting added event-production time for developing and implementing a security plan and managing security vendors, a requirement that fell outside the original proposal.

25. Do you have any unspent funds from the Grant? (required)

No

25.1. Please list the amount you did not use and explain why. Make sure to specify the currency. (required)

N/A

25.2. What are you planning to do with the underspent funds?

N/A

25.3. Please provide details of hope to spend these funds.

N/A

26. Are you in compliance with the terms outlined in the fund agreement? (required)

Yes

27. Are you in compliance with all applicable laws and regulations as outlined in the grant agreement? (required)

Yes

28. Are you in compliance with provisions of the United States Internal Revenue Code (“Code”), and with relevant tax laws and regulations restricting the use of the Funds as outlined in the grant agreement? In summary, this is to confirm that the funds were used in alignment with the WMF mission and for charitable/nonprofit/educational purposes.

Yes

29. If you have additional public recommendations or reflections that don’t fit into the above sections, please write them here. (optional)

The Bay Area Wikipedians User Group (BAWUG) played a key role in the success of the conference by spreading the word to Wikipedians throughout the region, connecting the dots to attract newcomers to the movement to the conference, and by contributing ideas, energy and creativity.
The support of the Internet Archive, including the relatively large number of employees who made time to attend, was also critical to the success of the 2026 conference.

While the organizing team has ideas about how to refine the format for the next WikiCredCon, survey respondents expressed strong satisfaction with WikiCredCon 2026:

"Just so lucky and grateful for this conference, thank you, thank you, thank you for everyone involved. If you come back, I promise Bay Area folks will have more dedicated resources for recommendations of places to go and things to do as well as more programming on the shoulders of the event if there is interest in it."
"Jake, Nevin, and Jenny did a fantastic job from start to finish. The lightning talks and other sessions were a good mix of high level, practical, and hands on collaboration. Great mix of people."
"The food!!! Also getting to know different people and projects that were in or connected to Wikimedia... Team did a great job, thank you!”

Plus, good food is an integral part of WikiCredCon:

Friday: the conference kicked off with a plenary dinner, with Chinese food from Mamahuhu.
Saturday: Arsicault Bakery breakfast, Tíos Chilos for Mexican lunch, and a Malaysian dinner from Damansara (with dessert from Uncle Tetsu).
Sunday: breakfast from Breadbelly and Good Luck Dim Sum, then Palestinian food from Mama Lamees for lunch.
So: Chinese, Mexican, Malaysian, dim sum, and Palestinian across the three days, with bakery breakfasts and a steady stream of snacks in between.