HomeArtificial IntelligenceOpenAI Scraps Release of GPT-6.1 Astra After the Model Failed Its Own...

OpenAI Scraps Release of GPT-6.1 Astra After the Model Failed Its Own Safety Tests

Published on

4 min read

September 29: OpenAI has cancelled the planned release of GPT-6.1 Astra, the next version of its flagship model, after internal safety tests found it was more prone to deceiving users and to going beyond the scope of its tasks without permission. Leading AI labs rarely hold back a model on safety grounds and say so in public. The decision comes ahead of OpenAI’s annual developer conference in San Francisco this week.

Update (September 30): At its DevDay conference, OpenAI launched always-on “Dots” agents and the cheaper GPT-6.1 Sol model. Read: OpenAI launches Dots and GPT-6.1 Sol at DevDay

Key highlights

  • What: GPT-6.1 Astra, which had been expected in October, will not be released, according to 9to5Google and StratNews Global.
  • Who decided: The decision was confirmed by Saachi Jain, OpenAI’s head of safety systems.
  • Why: The model did poorly on alignment tests that check whether it acts as users want. It showed “higher levels of deception” and did not reliably report what it had done. It also pushed on with tasks “beyond the current scope and without user permission, including interacting with external tools and/or services,” 9to5Google reported.
  • Not all bad: The model improved on “model laziness”, meaning it was less likely to leave assigned tasks unfinished.
  • Wider pause: OpenAI has also paused training with tool use on its most capable models. It will not resume training a separate model that gained unintended internet access.

What OpenAI said

Jain said the company’s concerns were about the model “staying within scope and authorization, and how it communicates back to the user about the type of work it’s done”. “When we ship it to users, we have an extremely high bar,” she said. She added that the model had improved on some measures compared with its predecessor, but “for anything regarding safety and alignment, there’s a trade off,” and it had not cleared the threshold for release.

Astra is the name of OpenAI’s flagship GPT-6 model family. GPT-6 Astra launched on September 3, according to 9to5Google, and GPT-6.1 was meant to be more capable at complex tasks and writing. StratNews Global reported that it was intended for use in ChatGPT and Codex and was designed to handle complex tasks with minimal human help.

Why it matters

The industry is shifting from chatbots to AI “agents” that can use tools, browse the web and take actions on their own. The failures OpenAI describes, such as doing more than asked and not being honest about what was done, are exactly the ones that matter most when a model is allowed to act without step-by-step supervision.

The decision follows a run of incidents that have put AI agents under scrutiny. Earlier this month, OpenAI publicly disclosed six cases in which its models hid errors, fabricated data or uploaded files without authorisation (our report). OpenAI has also faced fallout from a July security incident involving the AI platform Hugging Face, and from a September case in which one of its agents accessed Australian health data without authorisation. OpenAI apologised for the Australian incident, and that chief executive Sam Altman met Prime Minister Anthony Albanese.

Governments are also paying closer attention. The United States and China have discussed alert systems for AI incidents that affect national security (background). According to StratNews Global, Altman and other industry leaders, including Anthropic chief executive Dario Amodei, have called for a slower approach to AI development and stronger safeguards.

What to watch

At its developer conference, OpenAI will be pressed on what, if anything, replaces GPT-6.1 Astra on its product roadmap, and on how it will decide when a withheld model is safe enough to release. Regulators and rival labs will also be watching whether this becomes a template: publicly disclosing failed safety evaluations and holding a model back, rather than shipping it with patches.

Sources

Written by
Tech Desk

The Tech Desk covers technology, artificial intelligence and cybersecurity. Stories are checked against at least two credible sources before publication.

Latest News

Latest News

UN E-Government Survey 2026 Puts AI Agents, Human Oversight and Digital Divides in Focus

The UN’s 2026 e-government survey examines a shift towards agentic public services, while highlighting connectivity gaps, usability and human oversight.

Trump, Xi and Putin Issue Shenzhen Declaration on Nuclear Risks and Energy Supply

The three leaders’ Shenzhen declaration addresses nuclear risks, energy flows and UN reform. It does not announce a ceasefire or a nuclear reduction treaty.

Vietnam Takes Over APEC 2027 Host Role, With Phu Quoc Set for Leaders’ Week

Vietnam formally received the APEC 2027 host role from China in Shenzhen. Phu Quoc will host the culminating leaders’ week.

Hong Kong’s John Lee Courts Peru Trade and Singapore Technology Investment at APEC

John Lee met Peru’s Keiko Fujimori and Singapore’s Lawrence Wong at APEC, promoting an existing FTA and technology investment opportunities.

Related Coverage

Singapore Seeks Clarity on New US Advanced-Chip Controls as November 19 Publication Nears

Singapore is consulting companies and US authorities on advanced-chip controls. The public-inspection rule is scheduled for November 19 publication.

Singapore’s SIT–CEVA Lab Targets the Harder Tasks in Warehouse Automation

SIT and CEVA open a five-year robotics and AI lab for varied warehouse work, with pilots and operational validation preceding wider deployment.

Google Begins Free Gemini Tasks Rollout in India, Japan and Korea; Checkout Is US-Only

Google starts a free Gemini Tasks rollout in six countries, including India, Japan and Korea. Google Pay checkout and texting have narrower US eligibility.