Analysis in progress – figures are incomplete

The historical import of records for this mandate is still running. 0 of 14,225 records that can be analysed have been assessed so far (0.0%); 14,222 are waiting for analysis and 3 failed and will be retried. Totals and rankings will change as the remaining records are processed, so they do not yet represent the full mandate. How much has been processed

Who brings the culture war to Stormont?

MLA Culture Wars reads the Northern Ireland Assembly’s own record — written and oral questions, speeches, statements and motions — and asks an AI model, Jev, the same published questions about every one: is this Assembly business being used to stoke a sectarian, communal or identity-based row, or is it doing the job of scrutiny?

Every number links to the records behind it. Every classification shows its criteria and can be challenged. Mentioning the Irish language, the GAA, the Orange Order or flags is not enough to count — the use made of the issue is what matters.

Key figures

Source material: 14,225 Assembly records imported for this mandate, of which 0 of 14,225 analysable records (0.0%) have been assessed. Processing details.

Data last updated 20 October 2026.

The leaderboard

No records have been assessed yet, so there is no ranking to show. The leaderboard will appear as soon as the first records have been analysed.

Most recent qualifying activities

None yet.

How it works

1. Collect

Questions, Official Report contributions, motions and votes are imported from the Assembly’s Open Data service every few minutes, keeping the Assembly’s own identifiers and links.

2. Assess

Jev answers the same typed questions about every record — identity framing, antagonism, grievance, dog-whistles, escalation, scrutiny and stance — without being told who the MLA is or which party they belong to.

3. Check

Fixed thresholds turn answers into “qualifying”, “not qualifying” or “uncertain”. Uncertain records stay out of the totals until reviewed, and anyone can challenge a result.

Read the full methodology · See the exact questions Jev is asked