2026Q3 Minutes

Chair: Barbara Plank

Note-taker: Richard Johansson

Item 1 - Long-term submission system for COLING

Main topic and issues discussed

• COLING will join the ARR reviewing cycle for the first time and will share the October cycle with NAACL. The central question was how commitments should work when two conferences use the same review cycle.
• The main policy choice was between a hard commitment choice, in which authors must select either COLING or NAACL, and a preference-based model, in which authors can indicate first choice and acceptable backup options after seeing ARR reviews.
• Several participants argued that a hard choice creates a gambling problem for authors, especially students, because authors would have to infer from review scores whether to aim for the more selective or higher-prestige venue.
• COLING representatives were concerned that a hard-choice model could leave COLING with too few commitments if many authors choose NAACL, while NAACL representatives had reportedly preferred hard separation to preserve distinct conference identities and simplify management.

Additional issues raised

• The discussion distinguished ARR review from conference commitment. Participants emphasized that ARR should remain an independent review provider and that the commitment process should be coordinated by COLING, NAACL, ARR, and ACL rather than handled unilaterally by one conference.
• Acceptance-rate and prestige differences were discussed, as was the broader problem that ACL-family conferences increasingly have similar topical identities, making venue choice depend heavily on timing, location, and perceived prestige.
• A four-option preference scheme was considered: COLING only, NAACL only, COLING preferred with NAACL as alternative, and NAACL preferred with COLING as alternative. This was treated as a possible matching/routing problem rather than a review-process change.

Action points

• The executive committee supported exploring and recommending a preference-ordering option at commitment time, without treating that vote as a final decision for NAACL.
• A joint follow-up meeting was to be arranged with COLING, NAACL, ARR, and ACL executive representatives. Sina Zarrieß was asked to check whether Kevin and the NAACL team could attend the proposed meeting the following day at 11:00.
• The agreed direction was to avoid repeating the prior AACL/EMNLP submission-time split and to keep ARR review independent while coordinating the shared commitment cycle more carefully.

Item 3 - Preparation of AACL 2026

Main topic and issues discussed

• AACL-IJCNLP 2026 was reported as scheduled for November 6-10 in Hengqin, at a convention center able to accommodate more than 1,000 participants. The organizing committee was described as fully staffed, with general, program, local, tutorial, workshop, student, publication, ethics, and diversity/inclusion roles in place.
• Tutorials are being coordinated jointly with EMNLP; 59 tutorial submissions were received and about eight tutorials are expected to be accepted. Thirteen workshops were accepted on topics including multimodal and multilingual NLP, gender bias, alignment, mental-health applications, and knowledge graphs.
• The call for papers had gone out in June, with rolling deadlines through August, paper submissions due September 14, and camera-ready papers due in mid-October.
• The most serious concern was mutual exclusivity with EMNLP. In the May ARR cycle, about 17,000 valid submissions were recorded, with roughly 97.7% committed to EMNLP and only 1.8%, or 308 papers, committed to AACL. This was contrasted with roughly 1,200 submissions to the previous AACL-IJCNLP.

Additional issues raised

• AACL plans a phase-two commitment window after EMNLP notification so that prior ARR papers and EMNLP-rejected papers can still commit to AACL. This window is important because EMNLP notification precedes AACL notification by about 18 days.
• AACL governance and AFNLP collaboration were discussed. There will be no AACL 2027 organized by AACL; AFNLP will independently host an Asian NLP event in 2027, with AACL support. A joint AACL-AFNLP conference is planned for 2028.

Take aways and action points

• AACL 2026 will proceed with the planned phase-two commitment window after EMNLP notification.
• The organizers will continue sponsorship coordination with ACL sponsorship leadership, and continue keynote outreach.
• The nomination committee is to initiate the process for an AACL chair-elect as the current chair term concludes.
• No additional executive vote or formal resolution was recorded for this item.

Item 2 - Supporting conference PC chairs and PC submission policy

Main topic and issues discussed

• The discussion focused on the increasing difficulty of recruiting and retaining program chairs, and on whether PC chairs should be allowed to submit papers to the conferences they chair.
• The current policy environment was described as unclear and inconsistent. Some materials recommend that PC chairs not submit but still describe conflict-handling mechanisms, while a later executive resolution had barred conference chairs from submitting to their own conferences.
• PC-chair candidates and current chairs had raised practical concerns: travel funding may require a paper at the conference; the rule can affect students, postdocs, and collaborators; final-year PhD students can be harmed; and the policy may disproportionately affect people from institutions or countries with fewer resources.
• The proposal was to allow PC chairs to submit under independent conflict-of-interest handling, with strict recusal and with their papers ineligible for awards.

Additional issues raised

• Some participants worried about community perception if PC-chair papers are accepted, even when the process is fair. Others argued that ARR reduces some concern because review and commitment are more disentangled than in a conventional conference-only review system. There was also agreement that the COI policy should be communicated publicly so that the community understands the safeguards.
• Possible travel support for general chairs or program chairs who cannot otherwise attend was noted as a separate issue, not the main rationale for allowing submissions.

Resolutions

• The exec approved the suggested policy update. The supported policy is that PC chairs may submit under independent COI handling, while being strongly encouraged to avoid doing so when possible. PC chairs must have no role in decisions concerning their own papers, including acceptance, presentation format, and awards. Their papers are not to be eligible for awards.

Action points

• Barbara Plank will follow up with relevant chairs and update invitation materials, website text, policy pages, and templates so that the rules are clear before people accept roles.

Item 4 - Revenue status of ACL 2026 and ACL financial condition

Main topic and issues discussed

• ACL income has grown substantially in recent years, with registrations and membership increasing from about $4.2 million in 2022 to roughly $8 million in 2026. Total annual income is now roughly $9.5-$10 million, with sponsorship around $1.1 million, or about 14% of income.
• Expenses have also more than doubled and are now above $9 million. For 2026, ACL is aiming to be close to break-even overall, even though ACL 2026 itself is forecast to lose money.
• ACL has reserves on the order of $10 million, roughly 1.25 years of operating expenses. The treasurer emphasized that this is not an expendable surplus: it supports venue contracts, deposits, working capital, and ACL’s self-insurance against major disruptions.
• A major financial risk is venue scale. ACL and EMNLP have become too large for many hotels and convention centers, and the association increasingly faces expensive venues, limited negotiation leverage, and large hotel-block liabilities.

Additional issues raised

• EMNLP 2027 was used as an example of risk: the best candidate venue in the Americas outside the United States and Canada would require about $3.6 million in hotel-block guarantees, with additional location-specific weather risks.
• Membership income of roughly $800,000 no longer appears sufficient to cover journals, ARR, OpenReview, the ACL office, and other infrastructure. Raising membership fees and rethinking how shared infrastructure is funded were discussed.
• Staffing and volunteer management were raised as structural issues. Jenn is effectively managing a much larger organization and many volunteer roles by default; this does not scale and points to the need for clearer responsibility structures.
• The association’s investment approach has become more active, with a roughly 2% return on cash/CDs and attention to currency exposure.

Action points

• The executive committee should continue evaluating reserve needs, venue risk, self-insurance, hotel-block exposure, and infrastructure funding, including possible membership-fee changes.
• The staffing and volunteer-accountability concerns were carried forward into the later discussion on formal responsibilities for conference roles.

Item 7 - Use of LLMs and AI in scientific workflows

Main topic and issues discussed

• The Ethics Committee presented a draft statement on responsible AI use in research. The statement is built around the acronym STEWARD: Safety, Transparency, Evaluation, Watchfulness, Accountability, Reliability, and Data stewardship.
• The core principle is that researchers remain responsible for every scientific decision, observation, interpretation, and conclusion, regardless of whether AI tools are used. AI is treated as assistive technology and cannot claim authorship.
• The draft addresses the research lifecycle: ideation and literature review, experimental design, data handling, writing and translation, peer review, and post-publication responsibilities.
• For peer review, the draft emphasizes confidentiality: unpublished manuscripts should not be uploaded to external third-party AI services unless explicitly permitted, and confidential materials should be processed only with authorized systems.

Additional issues raised

• Several participants asked how the policy would be enforced and whether it would be tied to checklists, metadata, or other operational processes.
• The draft was encouraged to address post-publication corrections and retractions, and to be cautious about LLM-as-a-judge and AI-assisted annotation so the policy is not read as permission to use such systems blindly.
• Participants questioned the explicit allowance of AI-assisted hypothesis generation and emphasized the need to train new researchers in traditional scholarly practices, such as reading cited work.

Action points

• The Ethics Committee will revise the draft and coordinate with the publication ethics committee, the peer review committee, and other relevant bodies.
• There was support for concrete examples of acceptable and unacceptable practice, especially to distinguish AI assistance on individual steps from outsourcing an entire research project or paper to an AI system.
• The operational enforcement process remains to be developed, including how disclosures are collected and how violations are handled.
• The policy direction endorsed in discussion was that AI may support parts of the research process, but researchers must retain oversight, authorship responsibility, and accountability for the work.

Item 9 - Peer Review Committee report

Main topic and issues discussed

• The Peer Review Committee reported on review quality, review quantity, and submission quality. The immediate pressure point was the May ARR cycle, with about 17,000 valid submissions and only about 17% of submitting authors qualified to review. Around 40% of papers had no qualified reviewer among the authors.
• The committee proposed a service-credit or “coin” model. Papers consume review capacity, and authors earn credit by reviewing, serving as area chairs, or serving as senior area chairs etc. Papers whose authors collectively contribute enough service would be guaranteed review; other papers would enter a lottery based on remaining review capacity.
• The suggested equivalences were adjustable but included one paper being covered by service such as for example four reviews, eight meta-reviews by an area chair, or 16 meta-reviews by a senior area chair. The number of lottery papers reviewed in a cycle would be determined after the submission deadline based on actual capacity. The exact hyperparameters will need to be worked out.
• Open questions included what happens when promised reviews are not completed, whether excess credit can carry over, how to handle resubmissions, what exceptions should exist, and how to avoid harming newcomers.

Additional issues raised

• Reviewer eligibility criteria were debated. The current criterion was described as three ACL Anthology papers total, at least two of them main-conference or Findings papers. Participants noted metadata problems in OpenReview profiles and questioned whether the threshold is too strict.
• Alternative pathways to reviewer qualification were discussed, including reviewer training, senior recommendations, PhD-stage criteria, workshop reviewing, and credit for relevant publications in other AI/ML communities such as NeurIPS or CVPR.
• A second part of the report concerned hallucinated references. The committee wants a common policy defining rejectable cases such as nonexistent references, wrong DOI/link/arXiv identifiers, major author-list errors, or materially misrepresented sources, with milder cases receiving warnings.
• A third part concerned author responses and rebuttals. Current rebuttals were described as long, burdensome, and sometimes scientifically unhealthy because authors run rushed experiments. Eliminating rebuttal entirely was viewed as too drastic, but shortening or simplifying it and creating more reviewer-to-reviewer discussion received interest.

Take aways and action points

• There was general consensus to move forward with some version of the service-credit/lottery mechanism, while further developing details and inclusive pathways into reviewer eligibility.
• The broad idea will be presented at the business meeting, with clear examples showing how the model would work for authors. Luke Zettlemoyer offered to present it in person if helpful.
• The hallucinated-reference policy remains open for further development, including whether ACL can help authors check citations before submission rather than only detecting problems after acceptance.
• The rebuttal process will continue to be reconsidered, with attention to limiting rushed new experiments and improving reviewer discussion.

Item 10 - Preparation for ACL 2027

Main topic and issues discussed

• The ACL 2027 venue and conference dates have been determined. The location will be revealed at the ACL 2026 closing session.

Item 11 - Status of ACL 2026

Main topic and issues discussed

• Issues for ACL 2026: visa concerns and AI-generated papers.

Additional issues raised

• This item connects to later discussions on AI-assisted submissions, hallucinated references, and the need for clear policies and checks

Item 15 - AI-assisted fraudulent submissions and repeated conference spam

Main topic and issues discussed

• The discussion focused on hallucinated papers, hallucinated references, and repeated low-quality or thinly sliced submissions across ACL venues, especially workshops.
• Philipp Koehn summarized the need to clarify what behavior is acceptable and what should be considered unacceptable. Viviane Moreira described a dedicated checking system used for papers, including false positives such as websites being incorrectly flagged.
• AI detection had also been attempted for reviews, and anthology-related requests concerning hallucinated papers were mentioned as a recurring issue.
• The discussion then shifted to repeated spamming of conference venues by specific authors, including submissions of thinly sliced papers to many workshops.

Additional issues raised

• Participants noted that enforcement can be hard on students and that automatic checks might help. This connected with the broader theme that ACL needs prevention, education, and clear rules before applying strong penalties.
• The item overlapped with the Peer Review Committee discussion of hallucinated references and with later discussions of author-level or reviewer-level consequences for repeated misconduct.

Action points

• Philipp Koehn was asked whether he would present this issue at the business meeting.

Items 13/8/9 - ARR workflows, growth, reviewer involvement, and peer-review support

Main topic and issues discussed

• ARR editors are handling hundreds of papers per cycle. Although two half-time editorial assistants are in place and helpful, the workload remains far beyond current support. Additional operational workflow support and approximately a full-time software developer for unscheduled initiatives were discussed as possible needs.
• The group considered whether people who are not yet full reviewer-eligible could help with checklist checks or other narrow process tasks. This might both reduce operational load and train future reviewers, but participants cautioned that inexperienced volunteers may require rechecking and may not reduce burden unless tasks are tightly scoped.
• Checklist design was discussed, including the need for conditional logic so authors are not penalized for missing items that are only required in certain cases. OpenReview’s lack of conditional form logic was identified as a recurring limitation.

Additional issues raised

• Resubmission handling was discussed. Reviewers may miss the response letter because many download PDFs and work outside OpenReview. Participants considered making the response letter optionally visible to current reviewers when authors want it visible.
• Secondary reviewers were discussed. The primary reviewer would remain responsible, while junior collaborators or students could receive viewing access to observe the paper page and discussion before becoming official reviewers.
• The review period in the recent cycle was especially tight because of setup and eligibility delays. Participants noted that simply lengthening nominal deadlines may not help if reviewers wait until the end, but lower review loads would make deadlines more manageable.

Action points

• The optional visibility of resubmission response letters should be brought back to the relevant ARR/peer-review group for further consideration.
• The use of junior assistants or secondary reviewers should be explored only where responsibilities are narrow and the primary reviewer remains accountable.
• Further work is needed on workflow automation, OpenReview improvements, operational staffing, and the financing of review-capacity support.

Item 16 - Matters related to TACL

Main topic and issues discussed

• TACL is receiving increasing submissions, with more than 80 submissions in the July cycle. Average time to decision remains relatively stable because editors-in-chief and the editorial team are doing substantial prescreening and desk rejection.
• The main pressure is reviewer recruitment. TACL has difficulty securing reviewers when major ARR cycles absorb reviewer availability or when reviewers mark themselves unavailable because of ARR duties.
• TACL asked whether service as a TACL reviewer or action editor could be recognized in the ARR author form or broader community-reviewing duty framework.
• The discussion considered an ACL-wide reviewing bank that would recognize service across ARR, TACL, CL, and other ACL venues, while also noting the complexity of exchange rates, inflation control, and cross-venue accounting.

Additional issues raised

• The new ARR service-credit model might indirectly help TACL by reducing overloading of senior community members, but cross-venue service credit remains difficult to implement fairly.
• TACL is seeking two co-editors-in-chief starting soon and is also recruiting additional action editors, especially in newer areas such as agents.

Resolutions and action points

• Executive members were asked to help spread the call for two TACL co-editors-in-chief.
• TACL will continue adding action editors and expanding reviewer pools in emerging areas.

Item 6 - Matters about SIGs

Main topic and issues discussed

• The first SIG issue was a request to approve a new SIG in a clinical -related area. The proposed SIG had submitted a supporter list of about 81 people, including 39 ACL members, and had provided a constitution including the previously missing ethics section.
• The remainder of the item summarized a survey of SIGs. Of 31 SIGs, 21 responded. Most described themselves as active or very active, though some reported declining activity, changing community interests, and pressure from the rise of large language models.
• SIGs asked for greater visibility, clearer communication with ACL, more responsive financial management, stronger support for long-established SIGs, and clearer information about what support is available and how to request it.
• Long-standing SIGs raised questions about whether some events should be called conferences rather than workshops, how recurring events can be approved more easily, and how ACL can help with Scopus or other indexing.

Additional issues raised

• SIGs also asked for better tracking of SIG funds and sponsorship money, invited-speaker support, earlier workshop-allocation information, networking across SIGs, sharing of best practices, and sharing technical scripts that help move non-OpenReview or non-Softconf proceedings into the ACL Anthology.
• Participants emphasized that SIG workshops and SIG events should not be framed as second-class. They are a major part of ACL activity and help sustain subcommunities.

Resolutions

• The new SIG was approved.

Take aways and Action points

• Raffaella Bernardi will inform the organizers so they can announce it during ACL.
• SIGs should be highlighted more visibly in the business-meeting presentation, including a statement of their importance to the ACL community.
• ACL should improve SIG visibility and communication, consider website improvements, share best practices, clarify support processes, and point SIGs to people who can advise on indexing and publicity.
• Changes from workshop to conference naming should be handled case by case with executive oversight; established cases may be grandfathered, but new renamings should not be unilateral.

Item 17 - Matters related to CL

Main topic and issues discussed

• Computational Linguistics reported a significant increase in submissions, about 48% year over year, and was already close to 300 submissions by the beginning of the month. Desk-rejection time has been reduced.
• The journal’s latest impact factor is above 10 for the first time, and it is ranked first in linguistics. A publicity editor has been appointed, and new awards include a CL High Impact Paper Award and a CL Distinguished Reviewer Award. The dissertation award continues, with 34 nominations this year.
• A special issue on ethics of NLP and computational linguistics is planned, with a November 27 deadline and three guest editors. CL also has an arrangement allowing accepted CL papers to be presented at AAAI-related venues.
• The first policy issue concerned an undeclared extended CL submission related to a concurrently submitted ARR paper. Because the prior paper was not yet published and the overlap was not declared, CL declined the submission.

Additional issues raised

• The CL/ARR case raised the need for cross-venue duplicate or overlap checking across ARR, TACL, and CL. Issues include anonymity, confidentiality, different platforms, and who should own the process.
• The second policy issue was a complaint that a review was AI-generated was discussed.

Action points

• ARR, TACL, and CL should hold a joint discussion on practical duplicate/overlap checking, possibly starting with metadata exchange at the beginning of ARR cycles.
• CL will continue handling suspected AI-generated reviews manually and case by case for now. Adding another reviewer, as was done in the discussed case, was considered reasonable.

Item 5 - Promoting recognition of Findings papers

Main topic and issues discussed

• This item arose from an ARR editors-in-chief discussion about an author whose organization was unsure whether a Findings paper supported conference travel or counted as a conference paper.
• The group noted that ACL insiders understand Findings, but external institutions may not. No clear, easily findable web statement currently explains the status of Findings papers.
• Participants agreed that ACL should provide an authoritative statement that Findings papers are accepted ACL-family conference papers and that they have opportunities to be presented at the associated conference.
• The wording should avoid promising a specific presentation format, such as an oral presentation, unless that is guaranteed. “Presentation at an ACL event” or similar wording was preferred over potentially misleading oral/poster terminology.

Additional issues raised

• The item exposed a broader documentation problem: ACL policies, website pages, wiki pages, invitation materials, and role guidance are fragmented and sometimes outdated.
• Participants suggested clarifying categories such as main conference papers, Findings papers, TACL and CL presentations, SRW papers, and workshops, because authors and institutions often conflate these categories.
• Publicity and deadline visibility were also mentioned as related problems, especially when deadlines fall during vacation periods and authors do not know where to find authoritative information.

Resolutions and action points

• ACL should create or restore a clear Findings status statement on the ACL website.
• Acceptance letters and future calls for papers should include a sentence and link pointing authors and institutions to that authoritative statement.
• ACL needs a clearer owner or maintenance process for website, wiki, and policy pages so that policy changes are reflected consistently.

Item 14 - OpenReview support

Main topic and issues discussed

• OpenReview has made a formal request for substantially increased ACL support.
• Participants recognized that ACL and ARR make heavy use of OpenReview and request custom workflows, but also noted that ACL differs financially from some large machine-learning conferences, which may have more industry-heavy budgets and different revenue models.
• Possible alternatives were discussed. Returning to Softconf was viewed as infeasible at ACL scale. Forking and self-hosting OpenReview is technically possible.
• A more attractive direction may be to reduce custom requests to OpenReview, build more scripts and tools on top of the OpenReview API, and pay for clearly defined support or major case-by-case work.

Additional issues raised

• A plausible annual support range was discussed, with a request-as-is too high without stronger justification and a clearer support model.
• The financial context was emphasized: ACL is also facing recurring support costs for ARR staffing, journals, MIT Press, the Anthology, the ACL office, web and wiki maintenance, and other infrastructure. Each additional recurring $100,000 has budget consequences.
• Examples from other venues, such as improved action-chair emails built through scripts on top of OpenReview, were mentioned as evidence that some workflow improvements could be handled outside the core platform.

Resolutions and action points

• The executive committee did not accept the request as-is.
• Before responding, ACL should talk to ARR, the technical teams, and the RAVES team, and develop a clearer vision of what platform support and workflow tooling are actually needed.
• ACL should negotiate with OpenReview, likely within a lower funding range and with a defined level of support.

Item 12 - Business meeting introduction

Main topic and issues discussed

• The president had begun preparing the business-meeting slide deck and asked executive members to contribute directly.
• A proposed panel or short survey on slowing down or rethinking science in the age of LLMs was dropped because there was not enough time.
• The business meeting should preserve time for questions, ideally at least the final 15 minutes, and the slide deck may need to be cut further to achieve that.

Additional issues raised

• Expected short updates included Luke Zettlemoyer on the ARR system update and Philipp Koehn on PC issues and fraudulent papers. Other contributors were encouraged to appear briefly so the community can see who is responsible for the work, but updates should remain short.

Resolutions and action points

• Everyone contributing to the business meeting should add slides directly to the shared deck.
• The final deadline for slides is Sunday before the thank-you dinner, around 6:00 p.m.
• People who may need to answer questions should be present in the room for the business meeting.

Item 18 - Upcoming ACL events

Main topic and issues discussed

• The events update focused on venue planning and the challenge of finding locations large enough for ACL-scale meetings. Jenn had been venue-shopping and considering space, weather, visas, grants, costs, recommendations, and fit for ACL’s room pattern.
• EACL 2027 is mostly settled except for one remaining contract issue. ACL 2028 will be in Europe.
• The main constraint is not only plenary capacity but also the need for many breakout rooms, workshops, parallel sessions, poster space, and social-event capacity. Several otherwise attractive locations lack sufficient workshop space.
• For 2029, North America, and probably the United States, was described as the realistic region.

Additional issues raised

• The group discussed whether future growth might eventually require splitting across multiple locations, but participants worried that this could fragment the community or turn some sites into viewing parties.
• Virtual and hybrid options were discussed as partial responses to growth, including more panels, possible single-track elements, better video incentives, and better integration of online and in-person attendees. Fully integrated hybrid conferences remain difficult.
• Future China or Asia venues were discussed, especially given the large ACL community in China and the importance of future AACL planning there.

Action points

• Venue planning should continue Raffaella Bernardi will inform the organizers so they can announce it during ACL.
• Jenn will continue exploring North American options for 2029 and future Asia/China options for the appropriate cycle.
• Executive members suggested using business meetings, hallway conversations, or informal focus groups to collect venue ideas from the community.

Item 19 - Formal agreements of responsibilities for General Chairs, PC chairs, Workshop chairs, and related roles

Main topic and issues discussed

• The discussion addressed the fact that the general-chair role was intended to be an active management role, not a ceremonial title. General chairs are expected to recruit and supervise conference teams, track deadlines, follow up on problems, and ensure that committee members fulfill their roles.
• When responsibilities are not performed, Jenn becomes the de facto general chair, managing people, tracking tasks, and filling gaps on top of operational and business responsibilities.
• Participants agreed that ACL needs formal job descriptions for general chairs and other key roles, including PC chairs, workshop chairs, publication chairs, publicity chairs, and related positions. These should describe responsibilities, reporting structures, timelines, expected time commitments, and escalation paths.
• Documentation is currently outdated and fragmented. Past debriefs, checklists, and role knowledge may exist in email or Mattermost, but there is no organized, current source of truth. Project-management software and reusable checklists were suggested.

Additional issues raised

• The group discussed who should supervise general chairs. Options included the current President, VP, past President, chapter leadership, or a regional leader, depending on the conference. The key point was that the supervisor must not be Jenn and must be clearly accountable.
• There was extended discussion of aligning the VP-elect/President sequence with the geographic rotation of ACL conferences, so that the President or relevant regional leader has regional knowledge and ownership of the conference they supervise.
• The current VP-elect nomination slate created a transition problem because the identified candidates did not align neatly with the desired regional conference sequence.

Resolutions and action points

• Two principles were agreed. First, ACL should create written responsibilities and reporting structures for general chairs, PC chairs, workshop chairs, and other key positions; people should explicitly agree to these responsibilities when accepting the role.
• ACL should adopt a long-term alignment plan in which the President or appropriate regional leader is clearly responsible for the conference in their region, and the VP-elect/presidential sequence is aligned with conference rotation where possible.