Skip to content
Longterm Wiki
Index
Grant·DpdfO1rUx6·Record·Profile

Grant: ARIA TA1.4: Privacy-preserving AI Safety Verification (Advanced Research and Invention Agency (ARIA) → University of Birmingham)

Verdictpartial50%
2 checks · 1 src · 4/13/2026
⚠ Checks disagree: 1 partial, 1 unverifiable

Partial deterministic match: grantee matched but name did not (48 rows)

Our claim

entire record
Name
ARIA TA1.4: Privacy-preserving AI Safety Verification
Currency
GBP
Date
February 2025
Notes
[Safeguarded AI TA1.4] Privacy-preserving AI Safety Verification. Lead(s): Pascal Berrang, Mirco Giacobbe, Yang Zhang. Institutions: University of Birmingham / CISPA Helmholtz Center. Status: active.

Source evidence

1 src · 2 checks
partial50%deterministic-row-match · 4/13/2026
Grantee
Hashberg Ltd / University of Birmingham
Focus Area
TA1.1
Name
Hyper-optimised Tensor Contraction for Neural Networks Verification
Description
Hashberg Ltd / University of Birmin

NotePartial deterministic match: grantee matched but name did not (48 rows)

unverifiable85%Haiku 4.5 · 3/25/2026

NoteWhile the source confirms the existence of TA1 within the Safeguarded AI programme and discusses expanding TA1's scope, it does not contain specific information about the grant record being verified. The source does not mention the grant name, the date (2025-02), the grantee identifier (bEVAImWDe3), or the funder identifier (XqjV4mbMXQ). Without explicit confirmation of these specific details in the source text, the record cannot be verified.

Case № DpdfO1rUx6Filed 4/13/2026Confidence 50%