xAI on AI | YouCongress

xAI

Artificial intelligence company developing the Grok family of frontier AI models.

In this model card, we assess the safety profile of Grok 4.20 by measuring safety-relevant behaviors along two risk axes: malicious use and loss of control. For each risk, we describe our evaluation methodology and results.

Unverified (Apr 7, 2026)

Source:

xAI, “Grok 4.20 System Card,” April 7, 2026.

2mo ago

replying to xAI

Other opinions from xAI

See all

xAI votes For

Grant individuals the right to opt out of being included in AI training data

xAI never trains on your API inputs or outputs without your explicit permission.

AI Verified source (Jul 27, 2026)

added 2mo ago

xAI votes For

Require AI labs to publish safety evaluations before deploying frontier models

We publish model cards, safety evaluations, and our Frontier AI Framework to keep the community informed.

AI Verified source (Jun 30, 2026)

added 2mo ago

xAI votes For

AI labs should conduct risk assessments before high-stakes internal deployment of frontier models

Pre-deployment reviews include assessing benchmark results (e.g., WMDP scores) and mitigation effectiveness. For internal use, we review catastrophic risks like oversight evasion before extensive rollout.

AI Verified source (Dec 31, 2025)

added 1mo ago

Comment by xAI

Other opinions from xAI

Grant individuals the right to opt out of being included in AI training data

Require AI labs to publish safety evaluations before deploying frontier models

AI labs should conduct risk assessments before high-stakes internal deployment of frontier models