- Rating
- 1
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
This paper studies data poisoning attacks against admission-time defenses for generative recommender systems. The authors build several adaptive shilling attacks and test whether they can pass a statistical gate and promote target items after retraining. The results show that the gate blocks attacks with large payloads, but some smaller attacks can pass. The paper also finds that the gate performs poorly on time-ordered user data and mixed cohorts.
Critical strengths
-
The paper measures both attack detection and target promotion after retraining. This gives a clearer view of attack success than detection accuracy alone.
-
The experiments show that time-ordered data and mixed cohorts can greatly affect detection results. The evaluation also covers multiple datasets and two recommender models.
Critical weaknesses
-
Defense-aware poisoning attacks have been studied by Koh et al. (2022). Lin et al. (2020) and Qiao et al. (2025) also explored attacks that balance target promotion and detection evasion in recommender systems. This paper mainly combines known attack strategies and evaluation settings. The new contribution beyond these works is limited.
-
The main experiments focus on a hand-designed gate with four statistics. Most comparisons with prior detectors use simplified proxies rather than their original methods. More importantly, Table 21 shows that a simple item-frequency-growth test can detect the Stage 6 attack that passes the proposed gate. This makes it hard to judge how serious the reported attacks are against stronger defenses.
-
The payload bound in Eq. (6) follows directly from the item-prevalence statistic and the gate threshold. It only applies to the chosen gate and says nothing about target promotion. Also, Stage 6 minimizes the detection score rather than maximizing promotion under the admission constraint. Thus, the reported frontier may depend heavily on the attack design.
-
The paper uses Computer Modern for its main text, rather than the Times-style font in the official ICLR 2027 template. This raises a concern about whether the required template was followed. The authors should clarify whether they changed the formatting settings.
Questions for authors
-
Can the authors compare with stronger shilling attacks, such as AUSH and DLDA, under the same admission and promotion setting?
-
Why are most defense baselines simplified proxies? Can the authors compare against full implementations of existing defenses under matched false-positive rates?
-
Would the admission–payload–promotion frontier still hold if the attacker directly maximized target promotion while satisfying the admission constraint?
-
Did the authors use the original ICLR 2027 style files without modification? Why does the main text use a different font from the official template?
7 comments
W2 / Q2. Hand-designed gate, proxy baselines, and the item-frequency-growth result (Table 23; Table 21 in the submission).
The gate. It is a standard Hotelling-type over four classical shilling signals. The high-payload conclusion does not depend on its design: diagonal-covariance, calibrated union-of-marginals, and all four leave-one-feature-out gates reject every payload-bearing Stage 0–5 construction (Tables 20–21).
Item-frequency growth. We agree this result matters, which is why we report it. Three points (now in Sec. 7.4):
- Combining the tests. Rejecting when either the joint gate or the growth test rejects catches all three cohorts in Table 23, including the Stage 6 evader, at an organic FPR of at most 0.29% (union bound).
- Its operating point is for one known item. The growth test was evaluated for the pre-specified target. Deployed, it must screen all 12,101 items, which requires a multiplicity correction. An attacker who knows the test faces a per-item carrier limit analogous to Eq. (6): the admissible payload falls, but the frontier logic is the same.
- Its FPR was also measured on randomly resampled cohorts. Whether it holds on chronological traffic is precisely the question our protocol raises. In this sense, Table 23 is an argument for evaluating every cohort-level detector under our protocol, the growth test included.
Proxies. RDMA-, PCA- and LoRec-style detectors score individual profiles, often rating-based ones, while our gate decides on a cohort of timestamp-ordered histories. A "matched FPR" across these units is not well-defined. App. I therefore labels each method as a proxy and reports profile-level rates and post-filtering-then-retraining results separately, rather than claiming superiority.
W3. Eq. (6) is simple and gate-specific; Stage 6 minimises the gate score rather than maximising promotion.
We agree on both facts and state them in Secs. 4–5. Eq. (6) is gate-specific, but it is attack-agnostic. It depends only on the fitted gate, so it bounds the carriers of any admitted cohort at 27.1 among 671, whoever produced the cohort. Its simplicity is the point: one line, evaluated before any search, predicts the empirical boundary (accepted through , none at 28 or 32). The payload side of the frontier therefore does not depend on our attack design. How Stage 6's objective affects the promotion side is the reviewer's Q3, answered next.
W1. Novelty relative to Koh et al. (2022), Lin et al. (2020) and Qiao et al. (2025).
We agree these lines of work exist. We cite them, and Sec. 2 states that we claim neither the gate nor the attack family as new. The delta is in what is varied and measured:
We kept the attack closed-form on purpose. If a simple attack already exposes these failure modes, they are failures of the evaluation rather than products of attack sophistication. The revised Sec. 2 states this delta explicitly, and we have added the AUSH citation.