OpenAI chief executive Sam Altman in the picture.
The competition between OpenAI and Anthropic has entered another intense phase following the launch of OpenAI’s newest artificial intelligence system, GPT-6 Astra.
OpenAI is presenting Astra as a major advancement in AI capabilities, particularly in areas such as software engineering, scientific research, cybersecurity and professional work. The release comes as Anthropic continues to strengthen its own position with increasingly capable Claude models, setting the stage for another closely watched contest between the two AI companies.
Astra is designed to go beyond simply generating answers. OpenAI says the system is capable of handling complicated, multi-step assignments and interacting with software environments to complete tasks. This includes activities such as conducting research, writing and testing code, preparing documents and working with digital tools.
That shift toward more autonomous AI agents is becoming an important part of the competition. Instead of requiring users to guide an AI through every individual step, newer systems are being developed to carry out longer sequences of work with less direct supervision.
OpenAI’s latest model is also making headlines because of its reported performance in cybersecurity. The company says Astra has reached its highest “Critical” cybersecurity capability classification under its internal safety framework. The model can reportedly identify vulnerabilities and, under controlled conditions, develop methods for exploiting weaknesses in secure systems.
The advancement has simultaneously raised concerns about how increasingly powerful AI systems can be monitored and controlled. OpenAI has acknowledged that stronger capabilities create additional safety challenges and has emphasized the need for enhanced security measures, monitoring and containment procedures.
Astra’s introduction is particularly significant because Anthropic has emerged as one of OpenAI’s strongest competitors in the business AI market. Claude has gained attention among professional users, particularly for coding, research and other workplace applications. As a result, the latest OpenAI release is being viewed as an attempt to strengthen the company’s position in areas where Anthropic has been making progress.
OpenAI has made ambitious claims about Astra’s abilities, describing the system as a potential step toward artificial general intelligence. Company president Greg Brockman has characterized the model as a significant technological leap, although whether today’s systems truly qualify as AGI remains a matter of debate.
The model’s rollout is also being handled cautiously. Astra is initially being made available to a limited group of enterprise customers, with broader access expected to follow. Some of its most powerful cybersecurity capabilities are being restricted while OpenAI evaluates the associated risks.
The timing of the launch adds further significance to the rivalry. Both OpenAI and Anthropic are rapidly improving their models, with each company competing to attract developers, businesses and organizations seeking AI systems capable of performing increasingly sophisticated work.
For businesses, the contest could have practical consequences. More capable AI agents could reduce the time required for research, software development, document preparation and other knowledge-based tasks. At the same time, companies will need to consider reliability, security and human oversight before allowing autonomous systems to handle sensitive operations.
It remains too early to declare an outright winner between OpenAI and Anthropic. Benchmark results can differ depending on the type of task being measured, while real-world performance can vary from one application to another. OpenAI’s claims about Astra therefore still need to be assessed through broader independent testing and practical use.
Nevertheless, the arrival of Astra has once again placed the OpenAI-Anthropic rivalry at the center of the AI industry. The competition is no longer focused solely on which chatbot can provide the best response. Increasingly, the race is about which company can build AI capable of completing complex work, operating digital tools and functioning with greater independence.
As both companies continue pushing their technology forward, the latest OpenAI model represents another major development in a rapidly evolving AI race—one in which capability, commercial adoption and safety are becoming equally important measures of success.