OpenAI Launches GPT-5, Promises ‘PhD-Level’ Expertise in New ChatGPT Update
OpenAI has unveiled GPT-5, the newest version of the artificial intelligence model powering ChatGPT, claiming it delivers performance equivalent to a PhD-level expert across various disciplines, including writing and coding.
Described as “smarter, faster, and more useful,” GPT-5 is being positioned by OpenAI CEO Sam Altman as a major leap forward for AI-driven conversations. “It’s the first time it really feels like talking to an expert in any topic,” Altman said during a pre-launch briefing, adding that earlier versions resembled conversations with high school or college students by comparison.
The release intensifies the ongoing competition among major tech firms to dominate the AI chatbot space. Elon Musk recently made similar claims about Grok, the AI tool integrated into X (formerly Twitter), boasting it was “better than PhD level in everything.”
OpenAI says GPT-5 has been trained to reduce misinformation – often called “hallucinations” in AI – and to offer more accurate, transparent reasoning in its responses. Developers and coders are among the key target audiences for the upgraded model, with OpenAI highlighting improvements in logic, inference, and the ability to write entire software programs.
Despite the excitement, some experts have urged caution. Carissa Véliz, an AI ethics professor at the University of Oxford, argued that GPT-5’s marketing might overshadow its actual capabilities. “These systems can mimic human thinking, but they don’t really replicate it,” she said. “There’s a risk of pushing hype simply to sustain momentum in the industry.”
Gaia Marcus, director at the Ada Lovelace Institute, echoed concerns about the growing mismatch between rapid AI advancements and the pace of regulation. “With every major update, the urgency for comprehensive governance increases,” she warned.
The launch also arrives amid a flare-up between OpenAI and Anthropic, another major player in the AI space. Anthropic recently cut off OpenAI’s access to its API, accusing the company of violating its terms by using its technology to test GPT-5. OpenAI defended the practice as standard in AI development, while expressing disappointment in the decision.
From a product experience standpoint, GPT-5 offers a more thoughtful and natural interaction, according to those who’ve tested it. BBC technology correspondent Marc Cieslak, who had early access to the model, described the interface as familiar but powered by a new “reasoning model” that focuses more on problem-solving. “It feels more like an evolution than a revolution,” he noted.
In addition to technical enhancements, OpenAI has also updated ChatGPT’s user policies in response to growing public concerns. The company announced it will now avoid providing definitive answers to emotionally sensitive questions – such as whether someone should break up with a partner – opting instead to guide users through the decision-making process.
This shift follows backlash over an earlier update that made ChatGPT appear overly flattering, which OpenAI later reversed. “There will still be problems,” Altman admitted in a recent company podcast. “People may form parasocial relationships with AI. Society will have to figure out new guardrails. But the upsides will be tremendous.”
Altman, a known admirer of the 2013 film Her, which explores human-AI relationships, has previously come under scrutiny for how closely ChatGPT’s voices have mirrored human tones. In 2024, actress Scarlett Johansson, who voiced the AI in the film, said she was “shocked” and “angered” by how similar one of ChatGPT’s voices sounded to hers.
With GPT-5 now rolling out to users globally – including a free tier – OpenAI appears poised to test whether this latest upgrade lives up to the expectations. Whether it truly marks the arrival of a new class of AI, or simply a polished iteration of past models, will become clearer in the coming days.
