Frontier AI companies on Bletchley Declaration
Metadata
| Source Table | policy_stakeholders |
| Source ID | 8zsjplmvd4 |
| Source URL | www.gov.uk/government/publications/frontier-ai-safety-commitments-ai-seoul-summit-2024 |
| Parent | Bletchley Declaration |
| Children | — |
| Created | Mar 21, 2026, 4:21 AM |
| Updated | Mar 21, 2026, 3:13 PM |
| Synced | Mar 21, 2026, 3:13 PM |
Record Data
id | 8zsjplmvd4 |
policyEntityId | Bletchley Declaration(policy) |
stakeholderEntityId | — |
stakeholderDisplayName | Frontier AI companies |
position | support |
importance | high |
reason | Major AI labs including OpenAI, Google DeepMind, Anthropic, and Meta participated and agreed to pre-deployment safety testing |
source | www.gov.uk/government/publications/frontier-ai-safety-commitments-ai-seoul-summi… |
context | [ "Companies agreed to give UK AISI early access to frontier models for safety evaluation", "Commitments were voluntary and non-binding; enforcement purely reputational", "Seoul summit (May 2024) expanded to 16 company signatories with more specific commitments" ] |
Source Check Verdicts
Last checked: 4/9/2026
The record claims the stakeholder is 'Frontier AI companies (unknown)'. The source text confirms that frontier AI companies are stakeholders in the Frontier AI Safety Commitments policy, but the source does not provide specific names or identities of these companies—it only states that 'leading AI organisations' have agreed to the commitments. The '(unknown)' designation in the record appears intentional, reflecting the lack of specific company identification in the available source material. Since the source does not contradict the claim (it simply doesn't provide the specific identities), and the record itself marks them as unknown, this is unverifiable rather than contradicted.
Debug info
Thing ID: 8zsjplmvd4
Source Table: policy_stakeholders
Source ID: 8zsjplmvd4