Business applications Koshiro - stock.adobe.com Tiago Ventura, Computer Weekly Since its public debut in 2022, ChatGPT has swiftly reshaped the technological landscape and global economy. As the first widely accessible artificial intelligence (AI) tool to deliver near-human cognitive performance (from holding conversations and drafting emails to coding and conducting complex research), it has captured the attention of the general public and the business world. ChatGPT became the fastest-growing consumer application ever, amassing over 100 million active users in just two months, and its success sparked a global AI arms race , as governments and tech giants poured vast investments into AI infrastructure, vying to engineer the most advanced systems. Today, generative AI (GenAI) may be mainstream, but ChatGPT remains pre-eminent. It ranks among the top five most-visited websites worldwide and consistently outpaces other models in adoption and reach. Here, we tell you everything you need to know about it. What is ChatGPT? ChatGPT , short for Generative Pre-trained Transformer, is a conversational AI chatbot capable of understanding and generating human-like text in response to a user’s questions and prompts. The platform can be accessed through its website, mobile app and some third-party integrations. The AI is free to use at its full capability for a limited number of prompts. After this, users will have access to a less powerful version, or can upgrade to GPT Plus for single use, or GPT Team, which is designed for teams and businesses to collaborate using the tool. Who created ChatGPT? ChatGPT was created by OpenAI and released in November 2022, and has since become a benchmark for accessible, general-purpose AI applications in people’s personal and professional lives. OpenAI was founded in 2015 by a team including CEO Sam Altman, president Greg Brockman and scientist Wojciech Zaremba, along with Elon Musk, Ilya Sutskever and John Schulman, all of whom are no longer part of OpenAI. The company began as a not-for-profit research lab devoted to advancing AI for the common good of humanity. Over time, it spawned a for-profit subsidiary (OpenAI Global LLC) to attract capital investment, with Microsoft as a major backer. As of 2025, the company is valued at approximately $300bn , a valuation propelled by breakthrough models like ChatGPT. How does ChatGPT work? Any AI is a computer system designed to perform tasks that normally need human intelligence. At its core, ChatGPT belongs to the field of natural language processing (NLP), which equips machines with the ability to interpret and generate human language. This is a large language model (LLM) which learns from vast amounts of text, much of which is sourced from the internet and physical written work, and is fine-tuned through human interaction. The system uses this information and training to predict the next most probable word in a sentence. ChatGPT forms coherent responses to user prompts, drawing on patterns and structures observed during training, allowing it to become a generative AI that creates new content. What can ChatGPT do? ChatGPT handles an extraordinarily wide range of tasks due to its level of access to internet sources and training. It can turn complex technical explanations into accessible language, compare and analyse topics, draft formal documents, generate and refine creative writing, summarise content, write and debug code, and engage in casual conversation. Equally adept at planning, task assistance and adding needed context, it is being used across a wide range of private and public sector organisations . The answers will adopt a confident tone, and will normally end with an expanding question that aims to prompt you to continue to use the AI. Is ChatGPT ethical? Despite its benefits, ChatGPT raises ethical questions. Its reliance on existing internet content leads critics to question whether its outputs amount to plagiarism or unfair use of original work. Simultaneously, there is debate over whether its efficiency in helping to eradicate menial and time-consuming tasks may prompt job displacement in roles traditionally handled by humans. On top of this, due to the powerful computer chips needed to run ChatGPT, greater power and energy sources are required to maintain the datacentres used to meet the demand of AI. Computer Weekly has previously reported that datacentre operators may be rolling back their sustainability commitments or postponing deadlines for achieving net-zero goals due to the increasing push towards AI development and maintenance. In terms of regulations, there is currently no AI-specific legislation in the UK, as the government has said it aims to adopt a “pro-innovation” stance with guidelines rather than specific laws. On the other hand, the EU AI Act provides comprehensive regulation for the entire European Union. Who is using ChatGPT? In short, everyone. From individuals to major organisations and governments , ChatGPT has been widely integrated. Matt Weaver, head of solutions engineering for Europe at OpenAI, revealed to Computer Weekly that more than 600 million people use ChatGPT weekly, with the platform processing about one billion messages every day. In 2025, the UK government struck a partnership with OpenAI to deploy advanced AI models in security research and to explore AI’s role in public services. PwC stands out on the corporate front as OpenAI’s first reseller and largest enterprise user, deploying ChatGPT Enterprise to more than 100,000 staff across the UK and US. This is a business version of the system that mitigates privacy concerns around client data by preventing training on proprietary content. Similarly, OpenAI developed ChatGPT Edu to deploy AI system to students, faculty, researchers and campus operations that can reason across text and offer further data security. Are highly regulated or data-sensitive industries using ChatGPT? Thanks to the development of data-sensitive systems like ChatGPT Enterprise and GPT Edu, and application programming interfaces (APIs) that allow developers to integrate OpenAI’s language models into their own applications, major players in the business world are adopting the AI technology in their systems. In the world of banking, for example, BBVA has acquired more than 11,000 ChatGPT licences to help employees automate routine tasks in an attempt to save an average of 2.8 hours per week, per employee, and a collaboration between OpenAI and Santander will see the Spanish bank accelerate its use of AI and roll out ChatGPT Enterprise to 30,000 employees by the end of 2025. Meanwhile, Morgan Stanley Wealth Management uses ChatGPT to enhance how financial advisors access the firm’s knowledge base and respond to clients. Other companies such as Virgin Atlantic and Klarna use the AI to improve their customer support services by incorporating ChatGPT into their business practices. Is my data safe with ChatGPT? According to OpenAI’s privacy policy , ChatGPT does not use end-to-end encryption, and all chat histories are stored and may be used to train future models. Consequently, users are advised not to “share sensitive information, such as passwords, credit card numbers, or similar information, via the chat”. This means that anything you tell ChatGPT or any files uploaded onto the platform are kept by OpenAI. The platform also collects user email addresses, IP addresses, device identifiers, cookie information and location data, and may be required to share data with affiliates or government bodies, but not advertisers. Users retain the right to delete their data on request. For businesses seeking stronger protections , OpenAI’s Enterprise, Edu and Team offerings ensure that customer data is neither used for training purposes nor shared. “When ChatGPT Enterprise is adopted by an enterprise, we don’t train on their data for anything. It’s fully encrypted. It’s entirely private to the organisation,” says Weaver. Is ChatGPT accurate? Though often impressively accurate, ChatGPT can produce confident-sounding yet incorrect answers, known as AI hallucinations . These stem from data gaps, biased training inputs, or flawed model assumptions. Hallucinations can look and read like a correct and accurate answer, but are factually incorrect, and may be entirely fabricated. According to the Hughes Hallucination Evaluation Model (HHEM), GPT-5 has a hallucination rate of just 1.4%, outperforming rival systems such as Grok-4 and Gemini-2.5. OpenAI also reports that GPT-5 is 45% less prone to factual errors than its predecessor, GPT-4o, although hallucinations still occur in roughly one out of 10 routine tasks. What is ChatGPT-5? Launched on 7 August 2025, GPT-5 is the latest iteration of the OpenAI tool, which is available across all tiers: including Free, Plus, Pro, Team, Enterprise and Edu. Boasting improved accuracy, more compelling written answers and enhanced coding capabilities, including generating websites, games and apps from a single prompt, OpenAI describes it as “the smartest, fastest, most useful model yet”. GPT-5 features a “real-time router” that dynamically selects the optimal mode for each query based on its complexity and intent, preserving deep reasoning capacity for challenging requests. It also introduces user-selectable “personalities”, which include Cynic, Robot, Listener and Nerd. This allows users to fine-tune the tone and style of responses.

The company states that the new AI model is “smarter across the board, as reflected by its performance on academic and human-evaluated benchmarks, particularly in math, coding, visual perception, and health”. Timeline: Computer Weekly articles about ChatGPT 8 December 2023: How generative AI is put to work in business applications 8 February 2023: Is business ready for ChatGPT? 12 May 2023: The value and limitations of ChatGPT for businesses 31 August 2023: Thousands of Klarna staff access ChatGPT Enterprise 14 February 2024: Microsoft: Nation-state hackers are exploiting ChatGPT 11 June 2024: Apple adopts ChatGPT to put the ‘A’ in AI 27 August 2024: Government backs Danish version of ChatGPT 11 February 2025: Elon Musk capitalises on DeepSeek confusion to bid for OpenAI 17 February 2025: AI models explained: The benefits of open source AI models 27 March 2025: Microsoft’s ‘fraying relationship’ with OpenAI blamed for datacentre expansion plan rollback 31 March 2025: Microsoft restates commitment to OpenAI amid analyst note about datacentre expansion rollbacks 14 May 2025: BBVA creates ChatGPT Store and expands use of the GenAI tool 5 June 2025: An immersive experience is the future of AI 22 July 2025: How OpenAI is making ChatGPT public and private sector-ready 22 July 2025: UK government signs partnership with OpenAI 6 August 2025: OpenAI now offers open AI models, but CIOs need to assess the risk 8 August 2025: OpenAI closes gap to artificial general intelligence with GPT-5 13 August 2025: Santander will make AI training mandatory for all staff in 2026 15 August 2025: Autocratic UAE gets democratic artificial intelligence Read more on Business applications 30 of the best large language models in 2026 By: Ben Lutkevich Ministry of Justice’s OpenAI deal paves way to sovereign AI By: Cliff Saran What is ChatGPT Enterprise? Pricing, uses and features By: Alexander Gillis OpenAI targets India with datacentre push CIOs looking for ways to say yes to the iPad in the enterprise CIOs are looking for ways to say yes to the iPad in the enterprise, despite the technological and cultural challenges associated ... CIOs are looking for ways to say yes to the iPad in the enterprise, despite the technological and cultural challenges associated ... What CISOs should take from the Hugging Face-OpenAI incident Security experts say the lesson isn't to abandon sandboxing, but to strengthen the security controls around surrounding systems ... Security experts say the lesson isn't to abandon sandboxing, but to strengthen the security controls around surrounding systems ... Red agents vs. blue agents: How to make AI better at defense The agentic AI playing field was heavily tilted toward offense, so researchers began using red team agents to help teach their ... The agentic AI playing field was heavily tilted toward offense, so researchers began using red team agents to help teach their ... When AppSec scanners become a supply chain attack vector New research shows how security scanners embedded in the software supply chain can be attacked to serve as a foothold for ... New research shows how security scanners embedded in the software supply chain can be attacked to serve as a foothold for ... 5G Quiz - Can you speak 5G? Next-generation 5G wireless technology will offer faster speeds and increased capacity. Do you speak the language well enough to ... Next-generation 5G wireless technology will offer faster speeds and increased capacity. Do you speak the language well enough to ... 5G expansion is coming, but where will operators reap profits? In this recap of industry blogs, networking pundits examine the profit potential of 5G expansion, responsible use of AI and some ... In this recap of industry blogs, networking pundits examine the profit potential of 5G expansion, responsible use of AI and some ... 5G, cloud embolden network outsourcing and ultimate virtualization Could the cloud, 5G, small cell technology, BYOD and carrier-provisioned networks completely virtualize network infrastructure? ... Could the cloud, 5G, small cell technology, BYOD and carrier-provisioned networks completely virtualize network infrastructure? ... Distributed computing: The infrastructure shift AI demands The hyperscale era is ending. AI's energy and latency demands are driving infrastructure toward the edge -- closer to users, ... The hyperscale era is ending. AI's energy and latency demands are driving infrastructure toward the edge -- closer to users, ... Why Facebook and the NSA love graph databases Graph databases play six degrees of separation to find real connections. See how IT teams can use the database approach for ... Graph databases play six degrees of separation to find real connections. See how IT teams can use the database approach for ... DevOps and Agile IT save mainframe training from skills quagmire The problem isn't that new hires aren't familiar with the complex, custom daily tasks of mainframe ops. The problem is that ... The problem isn't that new hires aren't familiar with the complex, custom daily tasks of mainframe ops. The problem is that ... HR makes major strides toward improving employee engagement Cloud vs. legacy ERP systems: Tug of war intensifies for SMBs Aging legacy ERP systems at SMBs seem to be getting plenty of scrutiny these days. Heightened consumer demands, shifting ... Aging legacy ERP systems at SMBs seem to be getting plenty of scrutiny these days. Heightened consumer demands, shifting ... Develop smart AI in CRM strategies to win and keep customers Of the three words that comprise customer relationship management , one word binds the other two. As necessity and ... Of the three words that comprise customer relationship management , one word binds the other two. As necessity and ...