[Hotstorage-sc] [EXTERN] Input regarding AI generated submissions to HS '26

Wed Jul 15 15:31:48 EDT 2026

I believe AI-mostly-generated papers have a role (e.g., OpenAI recently
solved an Erdos conjecture) but they need to be treated separately, and we
need to get authors to cooperate. Here's a perhaps crazy proposal.
Conferences should create an AI-mostly track, limit the number of
acceptances of such papers (say, 20% of all accepted papers), ask authors
to mark their papers as AI mostly, use AI to assist in the review of such
papers, and ensure the proceedings show the paper as mostly AI (e.g., must
add "(AI paper)" to the title).

On Wed, Jul 15, 2026 at 11:41 AM André Brinkmann via Hotstorage-sc <
hotstorage-sc at fsl.cs.sunysb.edu> wrote:

> Hi Vasily,
>
> regarding your point 2), you raise a very important point, and I think we
> have all asked ourselves this question. After reviewing far too many
> AI-generated papers this year, my own answer is no. I do not believe such
> papers benefit a conference, because in many cases they have been written
> without being properly thought through. The main value of a conference,
> discussing with the authors and gaining insights beyond what the paper
> offers, is then most likely lost.
>
> The difficulty, of course, is that distinguishing AI-generated content
> from genuine research has become nearly impossible. So I cannot offer a
> real answer. What I can say is that when one set of authors submits ten
> papers including similar techniques, I am rather pessimistic that they are
> really interested in research, and I doubt I would get much out of a
> discussion with them.
>
> I look forward to discussing this with many of you in Prague, and I hope
> you have more optimistic answers than I do.
>
> Best regards,
>
> André
>
>
> On 15. Jul 2026, at 18:53, Vasily Tarasov <vtarasov at us.ibm.com> wrote:
>
> Lots of reasonable suggestions in the replies.
>
>
>    1. Do the 10 papers have evaluation section? Do the reviewers/chairs
>    think that the evaluation sections – i.e., experimental results – were
>    synthesized w/ AI (no actual experiments were run)?
>
>
>
>    2. One important question, IMHO, is how we view the following case. A
>    paper was *largely* generated with AI (far beyond polishing the text or
>    writing some sub sections). The system prototype was actually built (w/ AI
>    or not) and experiments were run for evaluation (AI took those results in
>    and generated eval section).  The authors did the due diligence reviewing
>    and revising the paper before submission. The paper turned out to satisfy
>    all requirements of a solid HotStorage paper (novel, clear, topic
>    appropriate, evaluation shows strong impact, etc.). Are we fine with
>    accepting such paper?
>    I think the answer is yes. (Don’t know if we want to add AI as
>    co-author – that’s what many code commits have nowadays 😉)
>
>
>
> Best,
>
> *Vasily Tarasov*
> Manager, Principal Research Scientist
> vtarasov at us.ibm.com
> *IBM Data Platform / GPU*
>
> *From:* Hotstorage-sc <hotstorage-sc-bounces at fsl.cs.sunysb.edu> *On
> Behalf Of *Sudarsun Kannan via Hotstorage-sc
> *Sent:* Wednesday, July 15, 2026 8:06 AM
> *To:* Gala Yadgar <gala at cs.technion.ac.il>; Ali Anwar <aanwar at umn.edu>
> *Cc:* Andre Brinkmann <andre.brinkmann at uni-saarland.de>;
> hotstorage-sc at hotstorage.org
> *Subject:* [EXTERNAL] Re: [Hotstorage-sc] Input regarding AI generated
> submissions to HS '26
>
> Hi all, I agree on previous comments and having some bounds on the number
> of papers per author, and maybe this could be mentioned in the future CFPs.
> Besides higher-level policies, some thoughts: Given that there is no clear
> consensus across
> Hi all,
> I agree on previous comments and having some bounds on the number of
> papers per author, and maybe this could be mentioned in the future CFPs.
> Besides higher-level policies, some thoughts:
> Given that there is no clear consensus across most of the conferences on
> what form of AI use is acceptable or not, we could come up with more
> precise description (and examples) for HS paper during one of the SC
> meetings.
> Regarding deciding on the 10 papers, unless all the 10 papers are solving
> the same problem or proposing similar design, a few questions we could ask
> are whether each of the submissions (1) advances storage research and
> systems research in general, (2) solves or proposes to solve a challenging
> and an important problem, and (3) helps realize the goals of the "hot"
> storage workshop and only accept papers that meet the bar.
>
> Best,
> Suda (Sudarsun)
> Associate Professor
> Rutgers CS
> https://people.cs.rutgers.edu/~sk2113/
> <https://urldefense.proofpoint.com/v2/url?u=https-3A__people.cs.rutgers.edu_-7Esk2113_&d=DwMGaQ&c=BSDicqBQBDjDI9RkVyTcHQ&r=9ACgIpHPMvCLl2qNt6CDPIeSyoMUMdgz6SufhRu4R5Q&m=AHwDRsTwKDFwAObY6Uqg2wkG6I0leKz_3bZz_WvOFVapMZ6M47UsrfQFbszh7r9Z&s=zx0H4MJiF1bTyH5B3KpFK_2vWmlQnm0zkFAEeAhTd1w&e=>
> ------------------------------
> *From:* Hotstorage-sc <hotstorage-sc-bounces at fsl.cs.sunysb.edu> on behalf
> of Ali Anwar via Hotstorage-sc <hotstorage-sc at fsl.cs.sunysb.edu>
> *Sent:* Wednesday, July 15, 2026 10:20 AM
> *To:* Gala Yadgar <gala at cs.technion.ac.il>
> *Cc:* Andre Brinkmann <andre.brinkmann at uni-saarland.de>;
> hotstorage-sc at hotstorage.org <hotstorage-sc at hotstorage.org>
> *Subject:* Re: [Hotstorage-sc] Input regarding AI generated submissions
> to HS '26
>
>
> I agree that if the submissions are substantially similar in content, we
> should reject all of them, whether on the basis of self-plagiarism or
> another appropriate policy ground.
>
> However, if the submissions are meaningfully different, then the approach
> Erez suggested earlier makes more sense. Even in that case, I support
> limiting the number of submissions per author. OSDI has also introduced a
> submission limit this year. Such a rule would discourage authors from
> attempting this type of exploit in the future.
>
> Reviewing workloads have increased significantly, and we should not ask
> reviewers to evaluate multiple AI-generated papers that differ only in
> minor ways. More importantly, we should develop a systematic method for
> identifying related cohorts of submissions, either with support from the
> submission platform or through another tool. This would allow us to apply
> the same policy consistently and fairly to everyone.
>
> Regards,
> -Ali
>
> On Wed, Jul 15, 2026 at 7:03 AM Gala Yadgar via Hotstorage-sc <
> hotstorage-sc at fsl.cs.sunysb.edu> wrote:
>
> I share Anrdé's sentiment.
>
> Submitting the same (or very similar) work as 10 different papers (or even
> two, for that matter) is abusing the system. AI makes it easier but is not
> necessarily the major problem in this case. I hope the PC can find an
> assertive way to send the message that this is unacceptable.
>
> If the scenario is as extreme as I understand it, then I think that
> accepting even one of those papers is problematic, because it would mean
> that the trick worked and that there was no penalty for the abuse.
>
> Best,
> Gala
>
> *From:* =?utf-8?q?Brinkmann=2C_Prof=2E_Dr=2E_Andr=C3=A9_via_Hotstorage-sc?=
> <hotstorage-sc at fsl.cs.sunysb.edu>
> *Sent:* Wednesday, July 15, 2026 13:35
> *To:* Angela Demke Brown <demke at cs.toronto.edu>
> *Cc:* Andre Brinkmann <andre.brinkmann at uni-saarland.de>;
> hotstorage-sc at hotstorage.org
> *Subject:* Re: [Hotstorage-sc] Input regarding AI generated submissions
> to HS '26
>
> Sie erhalten nicht häufig E-Mails von hotstorage-sc at fsl.cs.sunysb.edu. Erfahren
> Sie, warum dies wichtig ist
> <https://urldefense.proofpoint.com/v2/url?u=https-3A__aka.ms_LearnAboutSenderIdentification&d=DwMGaQ&c=BSDicqBQBDjDI9RkVyTcHQ&r=9ACgIpHPMvCLl2qNt6CDPIeSyoMUMdgz6SufhRu4R5Q&m=AHwDRsTwKDFwAObY6Uqg2wkG6I0leKz_3bZz_WvOFVapMZ6M47UsrfQFbszh7r9Z&s=6mcR0iqoiPSb7tHog7ji7_80GIX_zeWGDtMVN-gKQIU&e=>
> Dear All,
>
> Reviewing is not much fun at the moment, and I assume that nearly any
> submitted paper uses some kind of AI support ... The difference here, from
> my perspective, is that the authors have submitted so many papers with
> similar techniques. They are therefore speculating that, statistically,
> some of these papers will be accepted, which is an unfair advantage
> compared to other authors. I would therefore desk reject for
> "self-plagiarism".
>
> Best regards,
>
> André
>
>
>
> On 15. Jul 2026, at 03:43, Angela Demke Brown via Hotstorage-sc <
> hotstorage-sc at fsl.cs.sunysb.edu> wrote:
>
> We ran into something similar with SOSP. In our case it was 9 papers from
> one company/pair of authors and 6 papers from another company and pair of
> authors, which on investigation appeared to be the same authors using
> different first names.
>
> Given the lack of policy in the CFP on generative AI use, and the ACM
> authorship policies, we felt we couldn't do anything other than (a) ask the
> authors to explain their use of AI and (b) let the papers be rejected on
> the basis of their quality. The authors in our case claimed AI use only to
> polish their writing, which is generally allowed.
>
> I have no solutions to offer to HotStorage, unfortunately.
>
> All the best,
> Angela
>
> On Tue, Jul 14, 2026, 7:50 p.m. Ashvin Goel via Hotstorage-sc <
> hotstorage-sc at fsl.cs.sunysb.edu> wrote:
>
> The 10 papers were submitted by the same set of authors.
>
> Ashvin
>
> On Tue, Jul 14, 2026 at 8:28 PM Erez Zadok <ezk at fsl.cs.sunysb.edu> wrote:
>
> That is a hotly debated topic nowadays among ACM EiCs and multiple
> conferences.  We just had heated debates about this on the ACM ATC SC too.
> Different people are trying different things to varying degrees of success,
> from requiring disclosure by authors, to automated tools, etc.  No perfect
> solutions.  ACM seems to have given up on getting an AI-detection tools but
> are looking into cite-checking tools.
>
> ACM recently loosened their AI policy: they no longer prohibit it outright
> but ask authors to disclose AI use if wanted, then ask the reviewers to
> decide what, if, and how much AI use is ok or not ok.  IOW, ACM now shifted
> the burden onto reviewers.  Some are ok with this change, others less so.
>
> One problem with any AI detectors is that actual text can easily be
> flagged as AI-generated even if the authors use AI to spell check, polish
> the text, or ask AI to shrink the text size: these are considered valid
> uses of AI that shouldn’t result in rejection.
>
> That said, the easiest way to catch bad AI use is via citation checks.  At
> TOS, we now validate each citation of a submission before sending it off to
> be reviewed.  Even with search tools, that’s more work but we have no
> choice: recently we caught several papers with bad/hallucinated cites deep
> into the reviewing process (thus wasting reviewers’ precious time).  Any
> paper with obviously bad/hallucinated cites gets desk-rejected, but we also
> email the authors to see if they can explain themselves first.
>
> So my suggestions are:
>
> 1. Papers with obviously large portions of AI text should be rejected with
> minimal reviews.  If authors didn’t spend any significant effort to develop
> the paper, then our reviewers shouldn’t waste their time either.  “You get
> what you paid for” principle.
>
> 2. Papers with too many hallucinated cites should be rejected right away
> without review.  1-2 cites max may be considered a silly mistake that
> students sometimes make (I’d ask the authors), but any more should be
> rejected.
>
> 3. I don’t think we should limit the number of submissions per author.
> While rare, it’s possible to see 2-3 papers submitted by a subset of the
> same authors.  Now, when you say “a set of authors had submitted 10
> papers,” *how many* are in the set?!  Is it the same exact authors who
> submitted 10 separate papers?
>
> Thanks,
> Erez.
>
> > On Jul 14, 2026, at 2:18 PM, Ashvin Goel via Hotstorage-sc <
> hotstorage-sc at fsl.cs.sunysb.edu> wrote:
> >
> > Hi all,
> >
> > The HS '26 PC chairs (Young-ri and Bryan) are asking for input from us
> regarding a set of paper submissions that have likely been AI generated.
> Two reviewers expressed concerns to the chairs. They found several papers
> with similar concepts applied to different areas, following a similar
> structure, language (including hard to understand or strange words) and
> flaws. The chairs found that a set of authors had submitted 10 papers (13%
> of the 79 submissions). The reviewers are currently leaning negative on
> several of these papers, pointing out that the papers are difficult to
> understand, describe concepts vaguely, and use unfamiliar language. The
> positive-leaning papers will be discussed at the PC meeting.
> >
> > The chairs are looking for input about the best course of action to
> handle this case of suspected AI-generated papers. On the one hand, ACM
> allows the use of undisclosed AI. Thus, rejections should be grounded on
> the paper's technical contribution and novelty, rather than AI use, which
> is hard to detect accurately using current AI detection tools. On the other
> hand, such AI-generated papers will overwhelm reviewers in the future.
> >
> > Let me know if you have any specific suggestions regarding handling the
> current case with the 10 submissions. More broadly, we should discuss if HS
> should put in place any policies for AI generated papers. Here are some
> examples I have seen:
> >
> > 1. Using multiple detectors and letting reviewers know so they can decide
> > 2. Desk rejecting papers with hallucinated references
> > 3. Limiting number of submissions by authors
> >
> > Thanks
> > Ashvin
> > --
> > Hotstorage-sc mailing list
> > Hotstorage-sc at fsl.cs.sunysb.edu
> > https://linkprotect.cudasvc.com/url?a=https%3a%2f%2flists.fsl.cs.sunysb.edu%2fmailman%2flistinfo%2fhotstorage-sc&c=E,1,91WQUag8E2zfENfSQCUdIqUFzUsc2dwmj_dG6E_0zdAYmoIAMtNJGuoDt7S9H89VbLR7XNpnG9ULmTg0t2jSikOi9TaZ72-ui7II1Yi5Gz9IZxhs5A,,&typo=1
> <https://urldefense.proofpoint.com/v2/url?u=https-3A__linkprotect.cudasvc.com_url-3Fa-3Dhttps-253a-252f-252flists.fsl.cs.sunysb.edu-252fmailman-252flistinfo-252fhotstorage-2Dsc-26c-3DE-2C1-2Cz8DobtW1BkooL5a0UNH0Ce6rw30e6DGipeCv9AQMN7tsuC2XM5d49YemzXRXUt7-5FW7l9VJuVm9ocASktZsTHaM-2DUuclKlYotowsgGzwhcHIBze7dcUCBCIUP4A-2C-2C-26typo-3D1&d=DwMGaQ&c=BSDicqBQBDjDI9RkVyTcHQ&r=9ACgIpHPMvCLl2qNt6CDPIeSyoMUMdgz6SufhRu4R5Q&m=AHwDRsTwKDFwAObY6Uqg2wkG6I0leKz_3bZz_WvOFVapMZ6M47UsrfQFbszh7r9Z&s=v1_T5qi9LSTog-lceFHvhlJ1Cn7vfehFxBWubuwFS2M&e=>
>
> --
> Hotstorage-sc mailing list
> Hotstorage-sc at fsl.cs.sunysb.edu
> https://linkprotect.cudasvc.com/url?a=https%3a%2f%2flists.fsl.cs.sunysb.edu%2fmailman%2flistinfo%2fhotstorage-sc&c=E,1,nlmZlDMIVG-YFJEIr_6CFY3hI3LjsjwftL5FSAkAjKNiSwmMyMvb6oq08XvjRJjMQkUIoGeGnkJX1dBjPIZ1Q8_EqHDAEbweSEKxln8Gl3X-7Q,,&typo=1
> <https://urldefense.proofpoint.com/v2/url?u=https-3A__linkprotect.cudasvc.com_url-3Fa-3Dhttps-253a-252f-252flists.fsl.cs.sunysb.edu-252fmailman-252flistinfo-252fhotstorage-2Dsc-26c-3DE-2C1-2COKoPXtu-2DmnaObnaEK68Trt0G3VMUZwc93VDvfavS9cNBfYCht-2D8zuM58VHWnfOJdnjBoRMsQLMz7gLHKV0QQvDEiNH59-5F5rsUcssAKf8bcEuFzI6No1n-26typo-3D1&d=DwMGaQ&c=BSDicqBQBDjDI9RkVyTcHQ&r=9ACgIpHPMvCLl2qNt6CDPIeSyoMUMdgz6SufhRu4R5Q&m=AHwDRsTwKDFwAObY6Uqg2wkG6I0leKz_3bZz_WvOFVapMZ6M47UsrfQFbszh7r9Z&s=A6aZLUaVR5n3uRkHVy4jfrgS2SFBSbxp2bS3hIsCYW0&e=>
>
> --
> Hotstorage-sc mailing list
> Hotstorage-sc at fsl.cs.sunysb.edu
> https://linkprotect.cudasvc.com/url?a=https%3a%2f%2flists.fsl.cs.sunysb.edu%2fmailman%2flistinfo%2fhotstorage-sc&c=E,1,3lw7q9H3StQDaRNHlQTbwDpr4TO4mF8vVo-i8mwVkbfW3P_ztaLRQJK67ouFG4ZvqYJdTYZ_ydwPgsexGxcIvhx3C6eYNs8XWV5afV0kjw,,&typo=1
> <https://urldefense.proofpoint.com/v2/url?u=https-3A__linkprotect.cudasvc.com_url-3Fa-3Dhttps-253a-252f-252flists.fsl.cs.sunysb.edu-252fmailman-252flistinfo-252fhotstorage-2Dsc-26c-3DE-2C1-2CfjHwe0z08xZ16GBMBIKts4hz1-2DZBF7QclPPfzoixqTQGRTfupgTjbL3oaaKv3CLRphWVOHG6YJ5SDJeKfPRpXv9aNruWrCvS1dYGLLliuMi3tA5q2S4qKmRA7e4h-26typo-3D1-26ancr-5Fadd-3D1&d=DwMGaQ&c=BSDicqBQBDjDI9RkVyTcHQ&r=9ACgIpHPMvCLl2qNt6CDPIeSyoMUMdgz6SufhRu4R5Q&m=AHwDRsTwKDFwAObY6Uqg2wkG6I0leKz_3bZz_WvOFVapMZ6M47UsrfQFbszh7r9Z&s=rB-cmLors8Wvyp1IqopLSfSjMIRRdVggadxDAXaStAk&e=>
>
>
> --
> Hotstorage-sc mailing list
> Hotstorage-sc at fsl.cs.sunysb.edu
> https://linkprotect.cudasvc.com/url?a=https%3a%2f%2flists.fsl.cs.sunysb.edu%2fmailman%2flistinfo%2fhotstorage-sc&c=E,1,gr5auNENABq30J6uj4H-nZRqnU3_1gsAxJTfeiLlv_smX_jAEn_w4mTGLmJ6h66e_YwdMWUmWhIyLsHxICtsPWomMJetf41likdp5go_&typo=1
> <https://urldefense.proofpoint.com/v2/url?u=https-3A__linkprotect.cudasvc.com_url-3Fa-3Dhttps-253a-252f-252flists.fsl.cs.sunysb.edu-252fmailman-252flistinfo-252fhotstorage-2Dsc-26c-3DE-2C1-2ClAPnJvqR92iGPUK2fa-2Dz1YSNms9ejuxSDj1NQatyMtzm49i637UUy8cXEXSWMWQcl8IRsKy8LtNl98Ctl3z9yyezQDklLvem2RXGTViW-2DJb8-26typo-3D1&d=DwMGaQ&c=BSDicqBQBDjDI9RkVyTcHQ&r=9ACgIpHPMvCLl2qNt6CDPIeSyoMUMdgz6SufhRu4R5Q&m=AHwDRsTwKDFwAObY6Uqg2wkG6I0leKz_3bZz_WvOFVapMZ6M47UsrfQFbszh7r9Z&s=r6aQzuocX1srj_EP7NqEUxq7wRGJ4ec-IXmnEUbJYvU&e=>
>
>
> --
> Hotstorage-sc mailing list
> Hotstorage-sc at fsl.cs.sunysb.edu
> https://linkprotect.cudasvc.com/url?a=https%3a%2f%2flists.fsl.cs.sunysb.edu%2fmailman%2flistinfo%2fhotstorage-sc&c=E,1,ZRS8B-M2Y9SGb6Dzs84gq637La-qySXKTwWetKTWZXvVLehnQd-nkJ-jB099pdtv5Bezb5SKYQx3Z-jQaDPO9iIfxSsx4iLXTZQBXmSAehhfWQ,,&typo=1
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.fsl.cs.sunysb.edu/pipermail/hotstorage-sc/attachments/20260715/20b354dc/attachment-0001.htm>