OpenAI launches GPT-5, pushing for Enterprise AI dominance amid soaring industry investments
OpenAI raises the stakes in the AI race with its most capable model yet, promising enterprise-grade expertise and on-demand software creation.
- Aug 8, 2025,
- Updated Aug 8, 2025 8:32 AM IST
OpenAI has officially unveiled GPT-5, the latest iteration of its large language model powering ChatGPT, opening access to all 700 million users of the chatbot. The company is positioning this version as its most capable yet, with a focus on business applications alongside improvements for consumers.
The release comes during a period of unprecedented capital expenditure in the AI sector. Industry heavyweights Alphabet, Meta, Amazon and Microsoft, which backs OpenAI, are collectively set to invest nearly $400 billion this fiscal year to expand AI data centres and infrastructure.
OpenAI is reportedly in early talks to let employees sell shares at a $500 billion valuation, up from the current $300 billion. In this competitive climate, top AI researchers are commanding signing bonuses of up to $100 million.
The company is highlighting GPT-5’s ability to handle software development, health-related questions, finance and complex problem-solving. “GPT-5 is really the first time that I think one of our mainline models has felt like you can ask a legitimate expert, a PhD-level expert, anything,” said OpenAI CEO Sam Altman. “One of the coolest things it can do is write you good instantaneous software. This idea of software on demand is going to be one of the defining features of the GPT-5 era.”
During Thursday’s demonstration, OpenAI showcased “vibe coding,” where GPT-5 created functional software from natural language prompts. Early reviewers noted strong gains in coding, science and mathematics, though some suggested the leap from GPT-4 was smaller than in previous upgrades.
Despite the advancements, GPT-5 is not ready to replace human workers entirely. Altman acknowledged it still lacks the ability to learn autonomously. Dwarkesh Patel, on his AI podcast, compared the current approach to teaching a child an instrument by passing along written instructions from previous learners, without iterative practice.
Three years after ChatGPT’s debut, which marked a turning point for generative AI, the challenge for OpenAI has been scaling improvements. Constraints on data availability and the complexity of large-scale training have limited progress. GPT-5 incorporates “test-time compute,” a method that allows the model to dedicate extra processing power to complex queries, a capability now available to the public for the first time.
“We need to build a lot more infrastructure globally to have AI locally available in all these markets,” Altman said, signalling that the scale of investment needed remains vast.
For Unparalleled coverage of India's Businesses and Economy – Subscribe to Business Today Magazine
OpenAI has officially unveiled GPT-5, the latest iteration of its large language model powering ChatGPT, opening access to all 700 million users of the chatbot. The company is positioning this version as its most capable yet, with a focus on business applications alongside improvements for consumers.
The release comes during a period of unprecedented capital expenditure in the AI sector. Industry heavyweights Alphabet, Meta, Amazon and Microsoft, which backs OpenAI, are collectively set to invest nearly $400 billion this fiscal year to expand AI data centres and infrastructure.
OpenAI is reportedly in early talks to let employees sell shares at a $500 billion valuation, up from the current $300 billion. In this competitive climate, top AI researchers are commanding signing bonuses of up to $100 million.
The company is highlighting GPT-5’s ability to handle software development, health-related questions, finance and complex problem-solving. “GPT-5 is really the first time that I think one of our mainline models has felt like you can ask a legitimate expert, a PhD-level expert, anything,” said OpenAI CEO Sam Altman. “One of the coolest things it can do is write you good instantaneous software. This idea of software on demand is going to be one of the defining features of the GPT-5 era.”
During Thursday’s demonstration, OpenAI showcased “vibe coding,” where GPT-5 created functional software from natural language prompts. Early reviewers noted strong gains in coding, science and mathematics, though some suggested the leap from GPT-4 was smaller than in previous upgrades.
Despite the advancements, GPT-5 is not ready to replace human workers entirely. Altman acknowledged it still lacks the ability to learn autonomously. Dwarkesh Patel, on his AI podcast, compared the current approach to teaching a child an instrument by passing along written instructions from previous learners, without iterative practice.
Three years after ChatGPT’s debut, which marked a turning point for generative AI, the challenge for OpenAI has been scaling improvements. Constraints on data availability and the complexity of large-scale training have limited progress. GPT-5 incorporates “test-time compute,” a method that allows the model to dedicate extra processing power to complex queries, a capability now available to the public for the first time.
“We need to build a lot more infrastructure globally to have AI locally available in all these markets,” Altman said, signalling that the scale of investment needed remains vast.
For Unparalleled coverage of India's Businesses and Economy – Subscribe to Business Today Magazine