Provision record
OpenAI · OpenAI Safety Standards · View original document ↗

Superalignment Research Commitment

Low severity Medium confidence Explicit document language Common · 287 of 352 platforms
Stay ahead of the changes
Track OpenAI and get the diff the day its terms change.
Share 𝕏 Share in Share 🔒 PDF
Document Record

What it is

OpenAI states it has dedicated a team and 20% of its computing resources to researching how to ensure that AI systems far more capable than current models remain aligned with human values and oversight.

This analysis describes what OpenAI's agreement states, permits, or reserves. It does not constitute a legal determination about enforceability. Regulatory applicability and practical outcomes may vary by jurisdiction, enforcement context, and individual circumstances. Read our methodology

ConductAtlas Analysis

Why it matters (compliance & governance perspective)

This commitment describes how OpenAI allocates internal resources toward long-term safety research for AI systems that do not yet exist; it signals the organization's assessment of risk timelines and its stated prioritization of alignment research.

Interpretive note: The document states the 20% compute commitment but does not define how compute is measured, over what period the commitment applies, or what verification exists.

Clause Stability Stable

0
Changes
3
Months Monitored
May 12, 2026
First Seen
May 22, 2026
Last Seen
This clause type exists across 4261 other provisions on other platforms.

Consumer impact (what this means for users)

This provision describes a research investment commitment rather than a consumer-facing right or obligation; it does not directly affect current users' data, fees, or legal rights, but it represents OpenAI's stated position on the safety trajectory of systems that may power future products users access.

How other platforms handle this

Tinder Medium

TINDER ASSUMES NO RESPONSIBILITY FOR ANY CONTENT THAT YOU OR ANOTHER USER OR THIRD PARTY POSTS, SENDS, RECEIVES, AND/OR ACTS ON THROUGH OUR SERVICES, NOR DOES TINDER ASSUME ANY RESPONSIBILITY FOR THE IDENTITY, INTENTIONS...

Perplexity AI Medium

we do not warrant that Offering descriptions are accurate, complete, reliable, current, or error-free.

Skillshare Medium

Please note that these third parties are responsible for their own privacy practices.

See all platforms with this clause type →
▸ View Original Clause Language DOCUMENT RECORD
"
We believe that superintelligence could arrive within the decade. Solving the technical problems around superintelligence alignment is one of the most important tasks of our time. We've established a Superalignment team to focus on this challenge, committing 20% of our compute to solving it.

Excerpt from OpenAI's Safety Standards

ConductAtlas Analysis

Institutional analysis (regulatory & governance intelligence)

REGULATORY LANDSCAPE: Long-term AI alignment research of the type described here is not currently subject to specific mandatory regulatory frameworks in most jurisdictions, though it intersects with the EU AI Act's requirements for systemic risk …

Insight

Unlock the full institutional analysis

Enforcement risk, jurisdiction flags, contract triggers, and due diligence action items.

Applicable regulations

EU AI Act - High Risk Provisions
EU
FTC Act Section 5
United States Federal

Provision details

Document information
Document
OpenAI Safety Standards
Entity
OpenAI
Document last updated
May 12, 2026
Tracking information
First tracked
May 12, 2026
Last verified
May 12, 2026
Record ID
CA-P-011957
Document ID
CA-D-00822
Evidence Provenance
Source URL
Wayback Machine
Content hash (SHA-256)
46e71f573cc43a08729a6d0f09664a16c71e3f8e5fb577e6a1437e692885647e
Analysis generated
May 12, 2026 16:33 UTC
Methodology
Evidence
✓ Snapshot stored   ✓ Hash verified
Citation Record
Entity: OpenAI
Document: OpenAI Safety Standards
Record ID: CA-P-011957
Captured: 2026-05-12 16:33:49 UTC
SHA-256: 46e71f573cc43a08…
URL: https://conductatlas.com/platform/openai/openai-safety-standards/provision/CA-P-011957/superalignment-research-commitment/
Accessed: Sept. 8, 2026
Permanent archival reference. Stable identifier suitable for legal filings, compliance documentation, and research citation.
Classification
Severity
Low
Categories

Other risks in this policy

Get the research letter

Companies change their terms quietly. We read every version and catch what actually changed. One email a week on the changes that matter and what they mean.

Frequently Asked Questions

What does OpenAI's Superalignment Research Commitment clause do?

This commitment describes how OpenAI allocates internal resources toward long-term safety research for AI systems that do not yet exist; it signals the organization's assessment of risk timelines and its stated prioritization of alignment research.

How does this clause affect you?

This provision describes a research investment commitment rather than a consumer-facing right or obligation; it does not directly affect current users' data, fees, or legal rights, but it represents OpenAI's stated position on the safety trajectory of systems that may power future products users access.

How many platforms have this type of clause?

ConductAtlas has identified this type of provision across 287 platforms. See the full comparison.

Is ConductAtlas affiliated with OpenAI?

No. ConductAtlas is an independent monitoring service. We are not affiliated with, endorsed by, or sponsored by OpenAI.