Vulnerabilities > CVE-2022-36022 - Use of Insufficiently Random Values vulnerability in Eclipse Deeplearning4J

047910
CVSS 5.3 - MEDIUM
Attack vector
NETWORK
Attack complexity
LOW
Privileges required
NONE
Confidentiality impact
LOW
Integrity impact
NONE
Availability impact
NONE
network
low complexity
eclipse
CWE-330

Summary

Deeplearning4J is a suite of tools for deploying and training deep learning models using the JVM. Packages org.deeplearning4j:dl4j-examples and org.deeplearning4j:platform-tests through version 1.0.0-M2.1 may use some unclaimed S3 buckets in tests in examples. This is likely affect people who use some older NLP examples that reference an old S3 bucket. The problem has been patched. Users should upgrade to snapshots as Deeplearning4J plan to publish a release with the fix at a later date. As a workaround, download a word2vec google news vector from a new source using git lfs from here.

Common Weakness Enumeration (CWE)

Common Attack Pattern Enumeration and Classification (CAPEC)

  • Brute Force
    In this attack, some asset (information, functionality, identity, etc.) is protected by a finite secret value. The attacker attempts to gain access to this asset by using trial-and-error to exhaustively explore all the possible secret values in the hope of finding the secret (or a value that is functionally equivalent) that will unlock the asset. Examples of secrets can include, but are not limited to, passwords, encryption keys, database lookup keys, and initial values to one-way functions. The key factor in this attack is the attackers' ability to explore the possible secret space rapidly. This, in turn, is a function of the size of the secret space and the computational power the attacker is able to bring to bear on the problem. If the attacker has modest resources and the secret space is large, the challenge facing the attacker is intractable. While the defender cannot control the resources available to an attacker, they can control the size of the secret space. Creating a large secret space involves selecting one's secret from as large a field of equally likely alternative secrets as possible and ensuring that an attacker is unable to reduce the size of this field using available clues or cryptanalysis. Doing this is more difficult than it sounds since elimination of patterns (which, in turn, would provide an attacker clues that would help them reduce the space of potential secrets) is difficult to do using deterministic machines, such as computers. Assuming a finite secret space, a brute force attack will eventually succeed. The defender must rely on making sure that the time and resources necessary to do so will exceed the value of the information. For example, a secret space that will likely take hundreds of years to explore is likely safe from raw-brute force attacks.
  • Signature Spoofing by Key Recreation
    An attacker obtains an authoritative or reputable signer's private signature key by exploiting a cryptographic weakness in the signature algorithm or pseudorandom number generation and then uses this key to forge signatures from the original signer to mislead a victim into performing actions that benefit the attacker.
  • Session Credential Falsification through Prediction
    This attack targets predictable session ID in order to gain privileges. The attacker can predict the session ID used during a transaction to perform spoofing and session hijacking.