Culture

OpenAI scraps release of new model over safety concerns in internal testing

The Guardian Australia (Technology) Reuters

29 Sept 2026 09:02 AEST Summary

OpenAI Artificial intelligence US News Technology Sam Altman

From the report

OpenAI is scrapping the release of GPT-6.1 Astra, a next-generation ⁠AI model planned for an October debut, over safety concerns raised by researchers ⁠during internal testing.

How outlets headlined it

Each outlet's own headline. ~~Struck through~~: the loaded words our version leaves out. Plainest first.

The Guardian Australia (Technology) OpenAI scraps release of new model over safety concerns in internal testing plain
TechCrunch OpenAI reportedly ditches model over safety concerns plain
BBC News (Technology) OpenAI scraps rollout of new model over safety concerns plain
The West Australian OpenAI scraps new AI model release over safety concerns plain

Headline check

There is nothing in this headline a machine can check against the report: no figure, no name and no quotation.

Nothing was measured here, so nothing is claimed. How this is checked

This is a summary. The indicators are not. Every score on this page was measured over the full 247-word report, which is the only way the numbers mean anything.

How it is written

Niral Score 7.1 · calm mundaneloaded

10 more readings adjectives, repeated words, sentiment, sourcing and 1 more and 5 at zero

Highlighted in the text. Words that carry a fact and quoted words are not counted.

model4safety4ahead2company2concerns2external2human2internal2release2report2showed2tasks2testing2tools2unsafe2use2

Sentiment +0.0 hostilefavourable Sourcing 65 unattributednamed sources Mood 52.9 sorrowjoyful

5 read zero: political diction, generational, gender, jingoistic, religious

Political diction 0 left-codedright-coded Generational 0 harder on the oldharder on the young Gender 0 no gendered framingframing throughout Jingoistic 0 reports to a readerwrites from a camp Religious 0 not invokedovert

How the article is written, not whether it is correct. How each is measured.

Measured on The Guardian Australia (Technology)’s report as filed, which is all we may publish of this one.

How each outlet filed it

Also covered by The West Australian. We hold only their summary of it, not the report, so there is nothing of theirs to measure here.

Outlet	Niral Score	Adjectives	Sourcing	Sentiment	Happiness
The Guardian Australia (Technology)as they published this story	7.1	6	65	0.0	52.9
TechCrunch	10.1	7	55	0.4	55.9
BBC News (Technology)	7.1	9	77	-0.1	53.4

How did this read?

Comments

Comments are not open yet. They will be once there is someone to read every one before it appears. If something here is wrong, tell us — that we do read.

Report a problem with these scores or this rewrite — it goes straight into the public ledger.

The same event elsewhere

3 other outlets filed this story. The scoreboard above is what they did differently.

TechCrunch OpenAI reportedly ditches model over safety concerns Niral 10.1 · mostly calm political 0 BBC News (Technology) OpenAI scraps rollout of new model over safety concerns Niral 7.1 · calm political 0 The West Australian OpenAI scraps new AI model release over safety concerns Niral 2 · calm political 0