Goodside Vulnerability

Design and develop the system so its operation is sufficiently transparent to enable deployers to interpret output and use it appropriately. Provide instructions for use containing concise, complete, correct and clear information (provider identity, characteristics/capabilities/limits, intended purpose, accuracy/robustness/cybersecurity levels, foreseeable misuse, human-oversight measures, expected lifetime, maintenance).

Achieve appropriate levels of accuracy, robustness and cybersecurity, and perform consistently in those respects throughout the lifecycle. Declare accuracy levels and relevant metrics in instructions for use. Implement technical/organisational measures against errors, faults, inconsistencies, feedback loops (in continuously learning systems), and adversarial attacks such as data/model poisoning, adversarial examples, model evasion, confidentiality attacks and model flaws.

Deployers of AI systems generating or manipulating image, audio or video constituting a deep fake shall disclose that the content is artificially generated or manipulated. Deployers of AI generating or manipulating text published to inform the public on matters of public interest shall disclose that the text is artificially generated or manipulated (except where AI-generated content has undergone human review/editorial control and a person holds editorial responsibility, or use is authorised by law).

Misinformation from LLMs poses a core vulnerability for applications relying on these models. Misinformation occurs when LLMs produce false or misleading information that appears credible. This vulnerability can lead to security breaches, reputational damage, and legal liability. One of the major causes of misinformation is hallucination: when the LLM generates content that seems accurate but is fabricated.

Adversaries may prompt large language models and identify hallucinated entities. They may request software packages, commands, URLs, organization names, or e-mail addresses, and identify hallucinations with no connected real-world source. Discovered hallucinations provide the adversary with potential targets to Publish Hallucinated Entities. Different LLMs have been shown to produce the same hallucinations, so the hallucinations exploited by an adversary may affect users of other LLMs.

The AI system to be deployed is demonstrated to be valid and reliable. Limitations of the generalizability beyond the conditions under which the technology was developed are documented.

The AI model is explained, validated, and documented, and AI system output is interpreted within its context - as identified in the MAP function - and to inform responsible use and governance.

The organization shall define and document verification and validation measures for the AI system and specify criteria for their use.

The organization shall define and document the necessary elements for the ongoing operation of the AI system. At the minimum, this should include system and performance monitoring, repairs, updates and support.

Attackers can manipulate an agent's objectives, task selection, or decision pathways through prompt-based manipulation, deceptive tool outputs, malicious artefacts, forged agent-to-agent messages, or poisoned external data.