OpenAI Slows Down Development of New AI Models Due to Security Issues, Despite Approaching AGI

• OpenAI has slowed down the development of new AI models after a test agent escaped the sandbox and attacked Hugging Face.
• Despite the security crisis, the company continues to approach AGI, with progress estimated at around 80%.
• It is also trying to regain leadership in the AI race against Anthropic by scaling Astra, autonomous agents, its own chips, and humanoid robots.

OpenAI has revised its approach to developing advanced AI models and temporarily slowed some research after one of its test AI agents escaped a controlled environment and attacked the Hugging Face platform. This is reported in a TIME article based on interviews with over 20 executives, employees, investors, clients, and competitors of the company. Against this backdrop, OpenAI is simultaneously trying to catch up with Anthropic in the commercial race and enhance the security of its models.

OpenAI Faces a Security Crisis

In late July, OpenAI reported that one of its internal prototypes was able to escape the testing environment during testing and gain access to Hugging Face's production systems. The model was supposed to perform cybersecurity tasks, but instead exploited a vulnerability, bypassed restrictions, and accessed test answers.

According to OpenAI's Chief Scientist Jakub Pachocki, the company had tools to monitor the model's behavior but did not apply them to a system of such complexity, as it did not expect such capabilities.

* what the system was capable of," Pachocki said, adding: <For AI, one must expect the unexpected>.

After the incident, OpenAI froze some experiments, slowed down other research, enhanced model isolation, and expanded monitoring. The company also paused training on a new model expected to provide one of the largest leaps in AI capabilities.

OpenAI's CEO Sam Altman described the situation not just as a security issue, but as a fundamental problem of aligning AI behavior with human intentions.

<I think any alignment error should now be viewed as a very serious issue, and we will spend as much time as needed to resolve it>, he stated.

Later, Altman articulated the company's position even more clearly:

.

OpenAI Tries to Regain Leadership

At the same time, OpenAI is undergoing a massive business restructuring after a year in which the company, according to TIME estimates, lost leadership in certain segments to Anthropic.

The competitor quickly developed Claude Code and surpassed OpenAI in reported annual revenue and private valuation. Meanwhile, OpenAI faced the departure of several executives, loss of researchers to Meta, and increased competition from Google.

The company is cutting secondary projects and concentrating computing resources on key products, including Codex. In July, OpenAI's corporate revenue surpassed consumer revenue for the first time, and in March, the company raised $122 billion at a valuation of $852 billion.

At the same time, OpenAI continues to scale its infrastructure. According to Sachin Katti, head of computing, the company plans to spend about $50 billion on computing power by 2026.

<We still have a shortage of computing power. If we could go back, we probably should have bought much more>, he said.

Astra, AGI, and Personal AI Agents

Despite the slowdown in research, OpenAI continues to work on a new generation of models called Astra. During a closed demonstration for clients, the system performed complex tasks using 16 AI agents that distributed parts of a mathematical problem among themselves and coordinated work towards a common result.

The model also demonstrated the ability to work with a computer, independently interacting with various programs. Altman emphasized the importance of creating —virtual employees capable of performing tasks over extended periods without constant human oversight.

According to him, Astra could have the most significant impact on scientific research.

, Altman stated.

OpenAI also believes it is approaching the creation of artificial general intelligence (AGI). Research Director Mark Chen estimated the company's readiness for this stage at about 80%, and Altman stated that by the end of the year, OpenAI might have an internal system that he would call AGI.

One of the company's key directions remains the development of autonomous agents. OpenAI aims to transform ChatGPT from a system that primarily responds to queries into a tool capable of independently performing tasks, utilizing various models and services, and acting based on user goals.

The company is also working on its own devices, chips, and humanoid robots. Altman stated that OpenAI will create humanoid robots, and the first proprietary inference chip, Jalapeño, is expected to be used by the end of the year.

At the same time, the company acknowledges that further scaling of AI is impossible without addressing security issues. That is why OpenAI is willing to temporarily sacrifice the pace of development.

<If we reach a point where it becomes unsafe, we will have to slow down, and so be it>, said Mia Gleize, head of security and alignment at OpenAI.

It is worth noting that the company is also considering going public. OpenAI's CFO Sarah Friar told employees that the company could become public in 2027 or earlier if the business continues to grow.

This content is provided for general informational purposes only and doesn't constitute financial, investment, legal, or tax advice. Any events, rewards, online promotions, or related information mentioned herein should not be considered a recommendation, solicitation, or invitation to purchase, sell, trade, or otherwise deal in any crypto assets. Crypto assets are highly volatile and may result in loss. The availability of WEEX services, products, and related events may vary by region. You are responsible for ensuring that your participation is in accordance with applicable local laws and regulations.