OpenAI announced on Thursday the launch of its latest artificial intelligence model, GPT-5.5, which it says delivers improved performance in coding, computer use, and advanced research tasks.
According to CNBC, the release comes less than two months after the debut of GPT-5.4, highlighting the rapid pace of development driving the AI sector.
OpenAI President Greg Brockman said during a press conference that the model stands out for its ability to accomplish more with fewer instructions.
He explained that it can analyze ambiguous problems and determine precise next steps, adding that it helps lay the foundation for how humans will use computers in the future.
The company is racing to keep up with competitors, including Google and Anthropic, whose latest model, Claude Mythos Preview, has drawn attention on Wall Street.
In comments reported by TechCrunch, Brockman described GPT-5.5 as another step toward building a “super app," a multipurpose platform combining tools similar to ChatGPT, Codex, and an AI-powered browser into a single service for enterprise users. The concept has also been discussed by OpenAI cofounder Sam Altman.
The idea of an all-in-one AI platform is also central to Elon Musk, who has expressed ambitions to transform X into a similar “super app.”
OpenAI has been releasing models at an increasingly fast rate, with multiple updates introduced in recent months.
Jakub Pachocki noted that the company is seeing significant short-term progress and expects even more impactful improvements in the medium term.
According to CNBC, GPT-5.5 shows strong performance in data analysis, code generation and debugging, running software, web research, and creating documents and spreadsheets.
The company said the model does not cross the “critical” cybersecurity risk threshold but meets “high risk” classification standards, meaning it could amplify existing harmful pathways rather than create entirely new ones.
Mia Glaese stated that GPT-5.5 underwent extensive third-party safety testing, including cybersecurity and biological risk assessments, with enhanced safeguards developed over several months.
Concerns about AI-driven cybersecurity risks have grown across the tech industry and among government officials, especially after Anthropic introduced its Mythos model earlier this month with advanced vulnerability detection capabilities.
OpenAI confirmed that GPT-5.5 is being rolled out to paid subscribers, including Plus, Pro, Business, and Enterprise users, through ChatGPT and its coding assistant Codex, with API access expected to follow soon under additional safety measures.
Follow the latest news on Google News
