An aerospace manufacturer prioritizing accuracy and security wants to use generative AI. Which LLM adoption plan BEST aligns with its risk appetite?

The correct answer is A. Developing a private LLM to automate non-critical functions. A private LLM keeps all data, model weights, and inference on-premises or within the organization's controlled infrastructure, which is critical for an aerospace manufacturer handling sensitive, export-controlled, or proprietary data. Developing it for non-critical functions furt

AI Security Risk Management

Question

Options

ADeveloping a private LLM to automate non-critical functions
BContracting LLM access from a reputable third-party provider
CDeveloping a public LLM to automate critical functions
DPurchasing an LLM dataset on the open market

How the community answered

(36 responses)

A
47% (17)
B
31% (11)
C
17% (6)
D
6% (2)

Explanation

A private LLM keeps all data, model weights, and inference on-premises or within the organization's controlled infrastructure, which is critical for an aerospace manufacturer handling sensitive, export-controlled, or proprietary data. Developing it for non-critical functions further reduces risk by limiting blast radius if something goes wrong. Option B (third-party provider) requires sending data externally, violating the security posture. Option C (public LLM for critical functions) is the worst choice-public exposure and critical function automation together represent maximum risk. Option D (purchasing a dataset) doesn't provide an LLM at all; it's just data.

Topics

#LLM Deployment Models#AI Risk Management#Data Security#Organizational Risk Appetite

Community Discussion

No community discussion yet for this question.

Full AAISM Practice