ACM Publications Policy: Guidance for SIGCHI Venues

Acknowledgements: This post summarizes discussions amongst the SIGCHI Publications Committee (see https://sigchi.org/committees/ for members). Thanks to Pejman Mirza-Babaei (SIGCHI VP Publications), Sven Mayer (SIGCHI Pubs Committee), and Ran Zhou (SIGCHI Pubs Committee) for support in preparing this post.

TL;DR:

• All authors must have made substantial intellectual contributions to the paper, be aware that the manuscript has been submitted, and are held responsible and accountable for the work.
• All authors must be human beings and identified via PCS at the time of submission.
  – Pseudonyms are permitted, as is authorship after death, with permission of next of kin.
  – AI tools, communities, and non-humans (e.g., landscapes, animals) cannot be listed as authors, but can appear in the acknowledgements.
  – Adding or removing authors once the review process begins is prohibited.
• The use of generative AI tools (such as ChatGPT) is permitted, but all authors are responsible for the content created by these tools, and the use of the tools must be disclosed.
  – Authors are responsible for plagiarism, misrepresentation, fabrication or falsification of content and/or references by AI tools, and could be sanctioned with penalties, such as a publication ban.
  – AI tools may be safely used to copy-edit author-generated content, but caution is advised in using them for generating content.

What Can I Get from Reading this Post?

The Association for Computing Machinery (ACM) released an updated publication policy in April, 2023. This policy is essential reading for all chairs that intersect with publication (e.g., general, technical program, paper, late-breaking work, workshops, doctoral consortium, proceedings, publication) within our SIGCHI family of conferences; however, we also encourage all authors, reviewers, and associate chairs to familiarize themselves.

To assist conference and journal volunteers, the SIGCHI Publications Committee has discussed the policy and worked with the ACM to ensure accurate interpretation within the context of our typical SIGCHI processes. This blog post highlights relevant parts of the policy, draws attention to the new parts surrounding the use of generative artificial intelligence (AI) — including large language models (LLMs) — in publication, and provides sample content that can be included in call for papers (CFPs) that adhere to the policy.

Key Points of ACM Publications Policy

• This policy applies to all submitted, accepted, and published articles in all ACM publication venues.
• ACM journals, magazines, and conferences, and ICPS conferences shall reference this policy in Calls for Papers, Instructions for Authors, and other solicitations of submissions.
• Authors must be accountable for the work that was done and its presentation in a publication.

Criteria for Authorship

The policy now clearly describes what it means to be an author, and specifies the following key points:

• Authors must be an identifiable human being.
• Authors must have made substantial intellectual contributions to some components of the work described in the manuscript.
• Authors are aware the manuscript has been submitted for publication to ACM.
• Authors agree to be held responsible and accountable for any issues relating to the correctness or integrity of the Work and compliance with all related ACM Publications Policies.
• All authors need to certify that the Work submitted is original, that the listed authors are the creators of the Work.
• Note: All individuals who meet the above criteria should be listed as authors on the Work. The practices of gift authorship, guest authorship, ghost authorship, and purchased authorship are clear violations of the ACM Publications Policy and when proven may have severe consequences for those found to have participated in such practices.

The SIGCHI Publications Committee has discussed the criteria for authorship with our members, and are aware of the ways in which these criteria may support or prevent some desired authorship practices. Specifically, the policy does support publishing under a pseudonym or pen name, but also requires that the ACM has accurate contact information for the real identity of the pseudonymous author. Further, the policy specifies that authors who die or become incapacitated prior to publication can be listed as co-authors with permission of their estate or next of kin. The policy, however, prevents authors that are non-human (e.g., sentient landscape) or authors who are named as a collective (e.g., a community), rather than as individuals within a collective. Further, generative AI tools cannot be named as authors (see next section for further details), and human copy editors can be used, but should not be named as an author.

Because authors must be aware of the submission and accountable for the work presented, SIGCHI venues are advised to require that all authors are identified at the time of submission, with no additional authors being added or removed during the review process or at the time of submitting the final camera-ready documents for inclusion in the ACM Digital Library (DL). Requiring all authors to be identified upon submission of the work not only promotes integrity within peer review, but also helps to prevent gift authorship (i.e., when co-authorship is given as a reward, payment, or incentive to someone who did not contribute significantly to a Work), ghost authorship (i.e., when someone who did contribute significantly to a Work is concealed to hide a potential conflict of interest with reviewers, hide an author who drafts a Work on behalf of an industry backer, or in some other way deceive the reader about the authors involved), paper mills, or purchased authorship. These (and other) forms of unacceptable authorship practices can be seen in the FAQs for the authorship policy. SIGCHI venues can reject previously accepted papers if authors are added during the review process; however, venues with revise and resubmit processes can choose to allow adding authors at the time of submission of the revision to reflect intellectual contributions made during revision. It is important to note that all exceptions to authors being added after the point of submission are in the remit of paper chairs or technical program chairs — not steering committees or general chairs — and conflicts created by the addition of authors need to be managed by the paper chairs/editors and associate chairs/editors.

Please note that we are working on improving the PCS system for authorship notifications to better support collaborative paper authoring.

Authors who do not meet the criteria, or who are facing a ban on publication can be acknowledged at the end of the Work, before the bibliography.

Use of Generative AI Tools

Through significant consultation, the ACM has taken a position on the use of generative AI tools within research publication, and has issued the following guidance:

• Generative AI tools and technologies, such as ChatGPT, may not be listed as authors of an ACM published Work.
• The use of generative AI tools and technologies to create content is permitted but must be fully disclosed in the Work.
• Basic word processing systems are to be considered exceptions to this disclosure requirement.

The FAQs go into greater detail on questions of how to disclose the use of generative AI.

After significant discussion, the SIGCHI Publications Committee would like to raise attention to the requirement that while the use of generative AI tools and technologies to create content is permitted, authors will be held responsible for any plagiarized, misrepresented, or falsified content generated by an AI tool in ACM submissions. In particular, ChatGPT is well known to generate fake references that look real (for example it uses the real name of researchers active in the field), including generating the bibtex for these fake references. If used by authors to support a point made in their work, these fake references could be interpreted as falsified content to support claims, which is a clear violation of ACM authorship policy.

As mentioned earlier, all authors agree to be held responsible and accountable for any issues relating to the correctness or integrity of their submitted work and such policy violation may carry serious consequences, as noted in the ACM Penalties for Publication Violations Policy. Level III violations (i.e., when an action violates ACM Policy in a way that a reasonable person would have known that the conduct was inappropriate) result in a publication ban for one year and a letter to each author’s Dean, chair, or supervisor. The ACM Policy on Plagiarism, Misrepresentation, and Falsification outlines numerous ways in which authors must be responsible for their work, and should be careful with the use of generative AI tools, to avoid potential violation sanctions.

To be responsible, for SIGCHI venues, the use of generative AI tools are better served in modifying original content (e.g., shortening, copy editing) than in generating new content.

Inclusion in Calls for Papers

As mentioned in the policy itself, a link to the authorship policy must be included in all calls for papers. TPCs and other SIGCHI volunteers may also wish to summarize key points of the policy, as we have done in this blog post, to make it easy for authors.

Please send any questions about the policy to the VP Publications for SIGCHI: sigchi-vp-publications@acm.org.