Least-Privilege Calls to Amazon Web Services
Bibliographic record
Abstract
We address least-privilege in a particular context of public cloud computing: calls to Amazon Web Services (AWS) Application Programming Interfaces (APIs). AWS is, by far, the largest cloud provider, and therefore an important context in which to consider the fundamental security design principle of least-privilege, which states that a thread of execution should possess only those privileges it needs. There have been reports of over-privilege being a root cause of attacks against AWS cloud applications, and a least-privilege set for an API call is a necessary building-block in devising a least-privilege policy for a cloud application. We observe that accurate information on a least-privilege set for an invoker of a method to possess is simply not available for most such methods in AWS. We provide a meaningful characterization of least-privilege in this context. We then propose techniques to determine such sets, and discuss a black-box process we have devised and carried out to identify such sets for all 707 API methods we are able to invoke across five AWS services. We discuss a number of interesting discoveries we have made, some of which are surprising and some alarming, that we have reported to AWS. Our work has resulted in a database of least-privilege sets for API calls to AWS, which we make available publicly. Developers can consult our database when configuring security policies for their cloud applications, and we welcome contributors that augment our database. Also, we discuss example uses of our database via an assessment of two repositories and two full-fledged serverless applications that are available publicly and have policies published alongside. We observe that the vast majority of policies are over-privileged. Our work contributes constructively to securing cloud applications in the largest cloud provider.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.018 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.002 | 0.001 |
| Scholarly communication | 0.003 | 0.004 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.004 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".