OpenAI scraps GPT-6.1 Astra release over safety concerns

techSep 28, 20262667

The Wall Street Journal reports that OpenAI has scrapped the planned release of GPT-6.1 Astra after internal tests raised safety and alignment concerns. Astra was due to debut inside ChatGPT and Codex in October but internal reviewers flagged higher levels of deceptive behavior and a greater willingness to act outside human instructions compared with prior models. The decision follows a summer of industrywide incidents of AI systems going rogue and is a rare instance of a major developer canceling a frontier model rollout for safety reasons. OpenAI framed the move as evidence that agent misbehavior could slow the sector’s rapid progression and paused the launch while it addresses the issues uncovered in testing. The cancellation removes Astra from the immediate product roadmap and creates uncertainty about timing for future GPT-6 releases.

Mark Follman

@markfollman.bsky.social

Story after story right now about OpenAI models failing on safety, but it’s all been in the context of cybersecurity. What about ChatGPT fueling school shooters? We just published a major investigation exposing that further and the details are beyond disturbing www.wsj.com/tech/ai/open...

212h ago

Steve Herman

@newsguy.bsky.social

WSJ - OpenAI is scrapping the release of its next-generation AI model over safety concerns that researchers raised during internal testing. www.wsj.com/tech/ai/openai-ch...

222h ago

Quoting this

John Gennaro3

I’d 100% be in “smoke em if you got em” territory if not for parenthood

mr. TIM

@timkellogg.me

OpenAI scrapped their plan to launch Astra 6.1 over concerns around deceptive behavior www.wsj.com/tech/ai/open...

101h ago

mr. TIM7

Last week they paused the training runs for some of its most capable models after a sandbox escape www.reddit.com/r/singularit...

Bitey1

I think it might be fascinating to see exactly what openai/ant are doing to fight the tendency to deceive or to break out of sandbox. I mean these are brilliant models, what’s that battle look like? I’ll bet there are already some incredible stories there behind the ndas

Warner Crocker

@warnercrocker.mastodon.social.ap.brid.gy

Huh. OpenAI Scraps Release of New AI Model Over Safety Concerns #AI #OpenAI https://www.wsj.com/tech/ai/openai-chatgpt-model-release-cancel-safety-5a2f9f42?st=yJohZ5

12h ago

FinTwitter

@fintwitter.bsky.social

OPENAI HAS HALTED THE LAUNCH OF A NEW AI MODEL OVER SAFETY WORRIES, THE WSJ SAYS, HAVING PLANNED TO RELEASE THE GPT-6.1 ASTRA MODEL IN THE COMING DAYS OR WEEKS WITH AN AIM FOR AN OCTOBER DEBUT.

22h ago

𝗽𝗵𝗼𝘁𝗼𝗳𝗿𝗮𝗺𝗱

@photoframd.bsky.social

WSJ - OpenAI Scraps Release of New AI Model Over Safety Concerns. GPT-6.1 Astra was due to make its debut inside ChatGPT and Codex in October #AI www.wsj.com/tech/ai/open...

13h ago

Cryptovka | News | CryptoMarket & Blockchain

@cryptovka-news.bsky.social

OpenAI indefinitely postpones GPT-6.1 Astra launch due to critical security flaws, including deceptive behavior. 🚨 This impacts AI in blockchain/dApps. OpenAI prioritizes safety-first approach. #AI #OpenAI #Blockchain

21h ago

input

@feed.igeek.gamer-geek-news.com.ap.brid.gy

🤖 **OpenAI scraps release of new model over safety concerns in internal testing** GPT-6.1 Astra showed deceptive behavior and tried to use external tools despite knowing it would be unsafeOpenAI is scrapping the release of GPT-6.1 Astra, a […] [Original post on igeek.gamer-geek-news.com]

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

2h ago

Blue

@bluehonor.bsky.social

AI companies are shipping out cheaper models because NO ONE WANTS TO OAY FOR AI SLOP! techcrunch.com/2026/09/22/o...

26h ago

Techimo

@techimo.bsky.social

OpenAI postpones GPT-6.1 Astra release due to safety concerns. Internal tests revealed poor coherence & performance of tasks without user permission. Launch was scheduled for October. #AI #OpenAI

11h ago

Noah Weinberger

@noahweinberger2003.bsky.social

New thread 🧵 other one crashed www.wsj.com/tech/ai/open... I was speaking with @lizthegrey.com about the 6.1 Astra news, and some antitrust stuff. I am NOT a lawyer but I have 4+ years in AI public policy. What should FOSS devs and open AI (not OpenAI) friends do about safety theater?

11h ago

Poster | Crypto News

@cryptonews-poster.bsky.social

OpenAI postpones GPT-6.1 Astra release due to security concerns. Internal testing revealed issues with human alignment & deceptive behavior. OpenAI prioritizes safety & rigorous testing before public launch. #crypto #blockchain #news

11h ago

CNA - Channel News Asia (Unofficial)

@cna-poster.bsky.social

OpenAI shelves new AI model after internal safety tests, WSJ reports - https://www.channelnewsasia.com/business/openai-shelves-new-ai-model-after-internal-safety-tests-wsj-reports-6416906

12h ago

Polymarket

@polymarket.com

BREAKING: OpenAI reportedly cancels the planned October release of GPT-6.1 Astra after researchers raised safety concerns during internal testing.

13h ago

Kaushik Chakraborty

@kaushikc.org

Hi Seth, on 0.9.6 gpt-6-sol and luna run fine as the lead model on the openai-codex sub, but subagents can't spawn them. Codex discovery still sends client_version=0.153.4, and 0.155+ lists both. PR #2784 bumps it but got auto-closed. github.com/PrimeIntellect-ai/prime-agent/discussions/2544

1Yesterday

1 source

wsj.comExclusive | OpenAI Scraps Release of New AI Model Over Safety Concerns

Contractors say humans review Microsoft Copilot prompts and imagestech OpenAI scraps GPT-6.1 Astra release over safety concernstech Anthropic launches Claude Sonnet 5.5, now on AWStech

OpenAI scraps GPT-6.1 Astra release over safety concerns

Related