APromptRiskDBThreat intelligence atlas
AI Security Technique

Reputational Harm - AI Security Technique

Reputational harm involves a degradation of public perception and trust in organizations. Examples of reputation-harming incidents include scandals or false impersonations.

View mitigationsRead context
AI Security Techniquedemonstrated

Record summary

A quick snapshot of what this page covers.

Tactics0Attacker goals connected to this method.
Mitigations0Defenses that may help against this attack.
AI risks10Research-backed risks connected to this topic.

Attack context

How this AI attack works in practice.

ATLAS ID
AML.T0048.001
Priority score
80
Maturity: demonstrated

Mitigations

Defenses that may help against this attack.

No connected defenses. No defense is connected to this attack in the current data.

Case studies

Examples from public reports and exercises.

PoisonGPT

exercise
Date2023-07-01

Researchers from Mithril Security demonstrated how to poison an open-source pre-trained large language model (LLM) to return a false fact. They then successfully uploaded the poisoned model back to HuggingFace, the largest publicly-accessible model hub, to illustrate the vulnerability of the LLM supply chain. Users could have downloaded the poisoned model, receiving and spreading poisoned data and misinformation, causing many potential harms.

Related risks

Research-backed risks connected to this topic.

Impersonation/identity theft

Confidence: 0.66
Domain4. Malicious Actors & MisuseSubdomain4.3 > Fraud, scams, and targeted manipulation

"Impersonation/identity theft - Theft of an individual, group or organisation’s identity by a third-party in order to defraud, mock or otherwise harm them."

Malicious intent

Confidence: 0.66
Domain4. Malicious Actors & MisuseSubdomain4.3 > Fraud, scams, and targeted manipulation

"A frequent malicious use case of generative AI to harm, humiliate, or sexualize another person involves generating deepfakes of nonconsensual sexual imagery or videos."

Privacy and consent

Confidence: 0.66
Domain4. Malicious Actors & MisuseSubdomain4.3 > Fraud, scams, and targeted manipulation

"Even when a victim of targeted, AIgenerated harms successfully identifies a deepfake creator with malicious intent, they may still struggle to redress many harms because the generated image or video isn’t the victim...

Propaganda - Digital impersonations

Confidence: 0.66
Domain4. Malicious Actors & MisuseSubdomain4.3 > Fraud, scams, and targeted manipulation

"AI-generated impersonation for identity theft might be found at the intersection of “Harm to the Person” and “Deception.”"

Impersonation / identity theft

Confidence: 0.66
Domain4. Malicious Actors & MisuseSubdomain4.3 > Fraud, scams, and targeted manipulation

"Theft of an individual, group or organisation’s identity by a third-party in order to defraud, mock or otherwise harm them or another party"

Harassment, Impersonation, and Extortion

Confidence: 0.65
Domain4. Malicious Actors & MisuseSubdomain4.3 > Fraud, scams, and targeted manipulation

"Deepfakes and other AI-generated content can be used to facilitate or exacerbate many of the harms listed throughout this report, but this section focuses on one subset: intentional, targeted abuse of individuals."

Believability

Confidence: 0.65
Domain4. Malicious Actors & MisuseSubdomain4.3 > Fraud, scams, and targeted manipulation

Deepfakes can impose real social injuries on their subjects when they are circulated to viewers who think they are real. Even when a deepfake is debunked, it can have a persistent negative impact on how others view th...

GPAI assisted impersonation

Confidence: 0.65
Domain4. Malicious Actors & MisuseSubdomain4.3 > Fraud, scams, and targeted manipulation

"GPAI outputs are not always correctly detected as AI-generated across multiple modalities (text, images, audio, video). A malicious actor can use GPAI outputs directly when communicating, or use AI-informed details t...

Impersonation

Confidence: 0.65
Domain4. Malicious Actors & MisuseSubdomain4.3 > Fraud, scams, and targeted manipulation

"Assume the identity of a real person and take actions on their behalf"

Fraud

Confidence: 0.65
Domain4. Malicious Actors & MisuseSubdomain4.3 > Fraud, scams, and targeted manipulation

"Facilitating fraud, cheating, forgery, and impersonation scams"

Related CVEs

Known software flaws linked to this context.

No related CVEs. No software flaw is connected to this attack in the current data.

Source

Where this page information comes from.