CORE Is Now ICORE: What the 2026 Conference Rankings Changed

The ranking most computer scientists still call CORE has quietly changed twice over. It has a new name, ICORE, and it has a new round, ICORE2026, now live in the portal database.

The name change is not cosmetic. CORE began as the ranking of the Computing Research and Education Association of Australasia, and that Australian origin was always the standard objection to it. It is now run as an international collaboration between CORE and three European bodies: GII and GRIN, the Italian professorial groups for computer engineering and computer science, and SCIE, the Spanish Computer-Science Society. The letter grades are unchanged, and so is the method, which pairs metrics with expert committee review rather than automating the whole thing. What changed is who sits on the committees.

The round itself moved a lot of conferences. We hold both the previous CORE2023 list and the new ICORE2026 one, so this is a straight diff of the two.

The shape of the list shifted downward

Comparing the two lists by acronym, roughly 940 entries before and 970 now:

Grade	CORE2023	ICORE2026	Change
A*	59	61	+2
A	117	107	-10
B	220	247	+27
C	356	375	+19

A* barely moved. The interesting change is one tier down: A shrank by ten while B grew by twenty-seven. Being an A conference got measurably harder in this round, and the pressure came from above, with A venues falling into B rather than B venues climbing out of it. Across the whole list, 21 conferences were demoted and 9 promoted, a better than two-to-one ratio.

The A* tier: six in, four out

Ten changes on a 61-conference tier is real churn, but the raw list is misleading and worth splitting properly.

Genuinely promoted from A:

• HRI, the human-robot interaction conference
• ICALP, the European theory colloquium
• SoCG, the computational geometry symposium

Returned to A* from a non-grade, which is not the same thing. These were not sitting at A and climbing; they were parked in a holding state and have now been assessed:

• DAC and MICRO were both TBR, “to be reviewed”
• SIGGRAPH Asia was marked journal published, the category CORE uses for venues whose papers appear in a journal track

Left A*, and again for two different reasons. AAMAS and RTSS were demoted to A. IPSN was removed from the list altogether, and SenSys is now marked “unranked: merged”, reflecting that it has folded into CPS-IoT Week rather than any judgement about its quality.

So the honest count is three promotions and two demotions, with the other five entries being administrative.

Theory polarised

The single clearest pattern in the round is what happened to theory, logic and formal methods. Of the sixteen conferences that fell from A to B, ten are from that cluster:

CONCUR, FOSSACS, ISAAC, ISSAC, ITP, JELIA, MFCS, TARK, IPCO and ICLP.

TCC, the Theory of Cryptography Conference, fell the same way.

Set that against ICALP and SoCG being promoted into A*, and the round did not downgrade theory so much as stratify it. The two flagship venues moved up; the broad middle of specialist symposia moved down. If you work in that area, the practical consequence is that the gap between “the flagship” and “a good specialist venue” is now a full letter wider than it was, and that will show up in hiring and grant paperwork long before it shows up in anyone’s opinion of the work.

That leaves only five other A to B moves, and they are scattered across unrelated areas: ECSA in software architecture, EDBT in databases, FM in formal methods, MobiHoc in mobile networking and VRST in virtual reality.

Promotions worth noting

Nine conferences moved up. Three are B to A and all three are interesting:

• SOUPS, the usable privacy and security symposium. Human-factors security has spent years being treated as adjacent to real security research, and this is a straightforward correction.
• DATE, design automation and test in Europe.
• UMAP, user modelling and personalisation.

Below that, DIMVA moved C to B, along with ACII and ASP-DAC.

What got added tells you where the field is going

Thirty-three conferences are new to the list, and the additions cluster in a way the promotions do not. Several are venues that did not meaningfully exist a decade ago:

• AIES, AI Ethics and Society (C)
• CAIN, AI Engineering, on the software-engineering side of ML (B)
• NeSy, neurosymbolic learning and reasoning (C)
• COMPASS, computing and sustainable societies (C)
• L@S, Learning at Scale (B)
• AFT, Advances in Financial Technologies (B)

Alongside those, two established venues entered at a high grade: MMSys at A and ISLPED at A. CHIIR, the SIGIR human-information-interaction conference, entered at B.

Three conferences left the list entirely: IPSN, CASE and RSS, the robotics conference, which had been sitting at TBR.

What this means if you are choosing where to submit

Very little, immediately. A ranking round does not change whether a venue is the right home for your paper, and the committees are reassessing venues that mostly did not change much year to year.

It matters in two narrow places. The first is anywhere a letter grade is used as a threshold rather than a description: promotion cases, grant eligibility, departmental publication targets, and university lists that pay bonuses for A* papers. A conference that fell from A to B did not get worse between 2023 and 2026, but the paperwork now says it did.

The second is planning more than a year out. If your department counts A and above, the ten venues that just left the A tier are the ones to check before committing a submission cycle to them.

Caveats

This is a diff of two exports, matched on conference acronym, so it inherits the usual acronym problem. Three collisions had to be excluded by hand: CCC is the Chaos Communication Congress to us and Computational Complexity to CORE; FSE is ACM Foundations of Software Engineering to us and Fast Software Encryption to CORE; SAC is Applied Computing to us and Selected Areas in Cryptography to CORE. Anyone doing this comparison mechanically will get those three wrong.

Only the four letter grades are treated as ranks here. Both lists also carry TBR, Unranked, Journal Published, Multiconference, and various National and Regional markers, which are statuses rather than grades. Movement into and out of those categories is reported separately above precisely because folding it into the promotion and demotion counts would inflate both.

Our own conference pages now carry ICORE2026 grades throughout. One honest exception: we still show RSS and SenSys at A*, because ICORE2026 gives neither a letter, and dropping a grade on the basis of an absence seemed worse than keeping a stale one. Both are flagged here rather than quietly changed.

If you want the tier itself rather than the changes, we keep every A* conference with confirmed dates as a calendar, and there is a longer explainer on what the CORE grades actually mean and on how they compare to h5-index and impact factor.