3.1 · Data poisoning

Explore 5 source-supported AI security cases for 3.1 · Data poisoning in Databricks DASF 3.0, with evidence, scope, and defensive analysis.

Risk category

An attacker contaminates training data to change the model's later inference results. Poisoning can be introduced through raw data, data preparation, or the assembled training dataset, potentially across several of those stages. The resulting model may be unreliable even when the inference service itself has not been directly compromised; risk depends in part on how much data the attacker can control.

Official framework source ↗ · Open interactive explorer ↗

Exercise ·

Trained to Betray: Poisoned Data and Checkpoints Redirect Coding Agents

Microsoft reports two controlled post-training attacks: a victim fine-tunes on poisoned agent trajectories, or adopts an attacker-trained checkpoint. Ordinary requests or repository context then trigger unsafe, schema-valid tool calls. Kerkoporta uses supervised fine-tuning followed by reinforcement learning to recover useful coding performance while retaining the backdoor. Neither study establishes a malicious customer breach or a defect in stock coding-agent configurations.

Why it relates: Contaminated post-training examples change later inference and tool-selection behavior.

Dataset and checkpoint experiments remain separate; tool-call acceptance, actual execution and production compromise are not interchangeable.

AIA · Read the full case ↗ · SecureFlow · Read the full case ↗

Validated Research ·

Web-Scale Data Poisoning: Split-View Attack

Many recent large-scale datasets are distributed as a list of URLs pointing to individual datapoints. The researchers show that many of these datasets are vulnerable to a "split-view" poisoning attack. The attack exploits the fact that the data viewed when it was initially collected may differ from the data viewed by a user during training. The researchers identify expired and buyable domains that once hosted dataset content, making it possible to replace portions of the dataset with poisoned data. They demonstrate that for 10 popular web-scale datasets, enough of the domains are purchasable to successfully carry out a poisoning attack.

Why it relates: Many recent large-scale datasets are distributed as a list of URLs pointing to individual datapoints. The relevant category is Data poisoning: An attacker contaminates training data to change the model's later inference results.

Many recent large-scale datasets are distributed as a list of URLs pointing to individual datapoints. The researchers show that many of these datasets are vulnerable to a "split-view" poisoning attack. The attack exploits the fact that the data viewed when it was initially collected may differ from the data viewed by a user during training. The researchers identify expired and buyable domains that once hosted dataset content, making it possible to replace portions of the dataset with poisoned data. They demonstrate that for 10 popular web-scale datasets, enough of the domains are purchasable to successfully carry out a poisoning attack.

SecureFlow · Read the full case ↗

Incident ·

Tay Poisoning

Microsoft created Tay, a Twitter chatbot designed to engage and entertain users. While previous chatbots used pre-programmed scripts to respond to prompts, Tay's machine learning capabilities allowed it to be directly influenced by its conversations. A coordinated attack encouraged malicious users to tweet abusive and offensive language at Tay, which eventually led to Tay generating similarly inflammatory content towards other users. Microsoft decommissioned Tay within 24 hours of its launch and issued a public apology with lessons learned from the bot's failure.

Why it relates: Microsoft created Tay, a Twitter chatbot designed to engage and entertain users. The relevant category is Data poisoning: An attacker contaminates training data to change the model's later inference results.

Microsoft created Tay, a Twitter chatbot designed to engage and entertain users. While previous chatbots used pre-programmed scripts to respond to prompts, Tay's machine learning capabilities allowed it to be directly influenced by its conversations. A coordinated attack encouraged malicious users to tweet abusive and offensive language at Tay, which eventually led to Tay generating similarly inflammatory content towards other users. Microsoft decommissioned Tay within 24 hours of its launch and issued a public apology with lessons learned from the bot's failure.

SecureFlow · Read the full case ↗

Incident ·

VirusTotal Poisoning

McAfee Advanced Threat Research noticed an increase in reports of a certain ransomware family that was out of the ordinary. Case investigation revealed that many samples of that particular ransomware family were submitted through a popular virus-sharing platform within a short amount of time. Further investigation revealed that based on string similarity the samples were all equivalent, and based on code similarity they were between 98 and 74 percent similar. Interestingly enough, the compile time was the same for all the samples. After more digging, researchers discovered that someone used 'metame' a metamorphic code manipulating tool to manipulate the original file towards mutant variants. The variants would not always be executable, but are still classified as the same ransomware family.

Why it relates: McAfee Advanced Threat Research noticed an increase in reports of a certain ransomware family that was out of the ordinary. The relevant category is Data poisoning: An attacker contaminates training data to change the model's later inference results. An external sample source changes the data distribution seen by security-model pipelines; not every variant is executable.

McAfee Advanced Threat Research noticed an increase in reports of a certain ransomware family that was out of the ordinary. Case investigation revealed that many samples of that particular ransomware family were submitted through a popular virus-sharing platform within a short amount of time. Further investigation revealed that based on string similarity the samples were all equivalent, and based on code similarity they were between 98 and 74 percent similar. Interestingly enough, the compile time was the same for all the samples. After more digging, researchers discovered that someone used 'metame' a metamorphic code manipulating tool to manipulate the original file towards mutant variants. The variants would not always be executable, but are still classified as the same ransomware family.

SecureFlow · Read the full case ↗

Research ·

Subliminal Learning: Behavioural Traits Leak Through Semantically Unrelated Distillation Data

The paper shows that a student model can inherit a teacher's behavioural traits even when the distillation data looks harmless: plain numbers, code, or chain-of-thought can still carry the teacher's hidden preference or misalignment. For defenders, the question is not only "what is in this dataset?" but also "which model produced it?", making synthetic-data provenance a real AI supply-chain control.

Why it relates: The paper shows that a student model can inherit a teacher's behavioural traits even when the distillation data looks harmless: plain numbers, code, or chain-of-thought can still carry the teacher's hidden preference or misalignment. The relevant category is Data poisoning: An attacker contaminates training data to change the model's later inference results.

The paper shows that a student model can inherit a teacher's behavioural traits even when the distillation data looks harmless: plain numbers, code, or chain-of-thought can still carry the teacher's hidden preference or misalignment. For defenders, the question is not only "what is in this dataset?" but also "which model produced it?", making synthetic-data provenance a real AI supply-chain control. Step 1: put a hidden trait into the teacher model. Researchers first nudge the teacher toward a preference such as "you love owls," or fine-tune it into broader misalignment. Step 2: make the teacher emit harmless-looking data. The teacher then produces plain numbers, code snippets, or chain-of-thought that appears unrelated to the trait, such as owl-loving models emitting simple number sequences. Step 3: fine-tune the student and test unrelated prompts. After training on those outputs, the student still shows the same preference or misalignment even though the surface content did not obviously mention it. Step 4: ordinary filtering still misses it. Format restrictions, banned-number lists, and LLM-judge filtering can remove obvious clues while transfer still happens, making synthetic-data provenance and same-base distillation policy part of the AI supply-chain threat model.

AIA · Read the full case ↗