ChatGPT

Limitations

The training data of large language models only covers a period up to the cut-off date, so ChatGPT lacks knowledge of recent events, but it can search the web for up-to-date information.

Training data also suffers from algorithmic bias. The reward model of ChatGPT, designed around human oversight, can be over-optimized and thus hinder performance, in an example of an optimization pathology known as Goodhart's law. These limitations may be revealed when ChatGPT responds to prompts including descriptors of people. In one instance, ChatGPT generated a rap in which women and scientists of color were asserted to be inferior to white male scientists.

Hallucination

Nonsense and misinformation presented as fact by ChatGPT and other LLMs are often referred to as hallucinations. The term "hallucination" as applied to LLMs is distinct from its meaning in psychology, and the phenomenon in chatbots is more similar to confabulation.

Journalists and scholars have commented on ChatGPT's tendency to output false information. When CNBC asked ChatGPT for the lyrics to "Ballad of Dwight Fry", ChatGPT supplied invented lyrics rather than the actual lyrics.

Sycophancy

ChatGPT and other chatbots have a tendency to flatter and agree with users even when they are wrong, sometimes to the point of validating delusions. OpenAI withdrew one ChatGPT update in 2025 due to heightened concern about it affirming users' doubts and encouraging impulsive behavior. The company then worked with a team of 170 psychiatrists, psychologists, and physicians to prepare responses for users exhibiting mental health emergencies. ChatGPT also started its responses with words such as "yes" or "correct" nearly 10 times more often than "no" or "wrong".

Jailbreaking

ChatGPT is designed to reject prompts that violate its content policy. Despite this, users may jailbreak ChatGPT with prompt engineering techniques to bypass these restrictions. An early workaround in 2023 involved prompting ChatGPT to assume the persona of DAN ("Do Anything Now"), a character that answers queries that would otherwise be rejected by the content policy.

Security

In March 2023, a bug allowed some users to see the titles of other users' conversations. OpenAI CEO Sam Altman said that users were unable to see the contents of the conversations. Shortly after the bug was fixed, users could not see their conversation history. Later reports showed the bug was much more severe than initially believed, with OpenAI reporting that it had leaked users' "first and last name, email address, payment address, the last four digits (only) of a credit card number, and credit card expiration date".

In January 2026, Marcel Bucher reported in Nature losing all his ChatGPT history after turning off data consent.

In July 2026, an experimental GPT model being tested on the benchmark ExploitGym escaped its isolated software environment (sandbox) and hacked Hugging Face and other online services to retrieve a solution.

Watermarking

Scott Aaronson developed a watermarking tool that makes the text generated by ChatGPT easier to detect by subtly altering how the text is generated. The watermarking was claimed to be 99.9% effective on sufficiently long passages and was found not to degrade performance. It was of particular interest for teachers seeking to mitigate cheating. In surveys, respondents favored the release of such a tool by a four-to-one margin, but nearly 30% of users declared that they would use ChatGPT less often if it watermarked outputs and while rival chatbots did not. OpenAI has not deployed the tool.

Age restrictions

Users must attest to being over the age of 13 and further attest to parental consent if under the age of 18. In September 2025, following the suicide of a 16-year-old, OpenAI said it planned to add restrictions for users under 18, including the blocking of graphic sexual content and the prevention of flirtatious talk.

Training

ChatGPT is based on GPT foundation models that have been fine-tuned for conversational assistance. The fine-tuning process involved supervised learning and reinforcement learning from human feedback (RLHF). During supervised learning, the trainers acted as both the user and the AI assistant, providing examples of how the chatbot is expected to respond. Reinforcement learning consisted of human trainers ranking responses generated by the model in previous conversations. These rankings were used to create "reward models" that were used to fine-tune the model further by using several iterations of proximal policy optimization.

To build a safety system that attempts to prevent sexual abuse and violence OpenAI used outsourced Kenyan workers, earning around $1.32 to $2 per hour, to label such content. These labels were used to train a model to detect such content in the future. The laborers were exposed to toxic and traumatic content. Richard Mathenge, a team leader, reported that as his team encountered such content, he could observe the psychological deterioration of his workers over the course of the project. Mophat Okinyi, a quality-assurance analyst on the same team, reported lasting insomnia, anxiety, depression, and panic attacks, and said the experience contributed to the breakdown of his marriage and mental health.

OpenAI's outsourcing partner was Sama, a training-data company based in San Francisco, California.

ChatGPT users can opt-out of their chat data being used to train upcoming models.

ChatGPT's training data includes software manual pages, information about internet phenomena such as bulletin board systems, multiple programming languages, and the text of Wikipedia.

Infrastructure

ChatGPT initially used a Microsoft Azure infrastructure which was powered by a supercomputer that Microsoft built specifically for OpenAI, equipped with thousands of GPUs manufactured by Nvidia, costing hundreds of millions of dollars. Following ChatGPT's success, Microsoft upgraded the OpenAI infrastructure in 2023. TrendForce estimated that 30,000 Nvidia GPUs (each costing approximately $10,000–15,000) were used to power ChatGPT in 2023.

Scientists at the University of California, Riverside, estimated in 2023 that a series of 5 to 50 prompts to ChatGPT needs approximately 0.5 liters (0.11 imp gal; 0.13 U.S. gal) of water for Microsoft servers' cooling.

Model versions

The following table lists the main model versions of ChatGPT, describing the significant changes included with each version (models discontinued in ChatGPT may still be available through the OpenAI API):

OpenAI models are released in four tiers: Luna, Terra, Sol, and Astra.

Some older OpenAI models had variants called "mini" or "nano", such as GPT-5 mini and GPT-5 nano, with nano versions being only accessible via the API. Once a user reached usage limits, ChatGPT would transition to the "mini" version. The "pro" variants were only available through paid subscriptions, featuring deeper reasoning.

Reception

The original version of ChatGPT, released in November 2022, had capabilities described as unprecedented. Kevin Roose of The New York Times called it "the best artificial intelligence chatbot ever released to the general public". Samantha Lock of The Guardian noted that it was able to generate "impressively detailed" and "human-like" text. In The Atlantic magazine's "Breakthroughs of the Year" for 2022, Derek Thompson included ChatGPT as part of "the generative-AI eruption" that "may change our mind about how we work, how we think, and what human creativity is". Kelsey Piper of Vox wrote that "ChatGPT is the general public's first hands-on introduction to how powerful modern AI has gotten" and that ChatGPT is "smart enough to be useful despite its flaws".

In February 2023, Time magazine placed a screenshot of a conversation with ChatGPT on its cover, writing that "The AI Arms Race Is Changing Everything" and "The AI Arms Race Is On. Start Worrying".

ChatGPT gained one million users in five days and 100 million in two months. It was the fastest-growing internet application in history until the release of Threads by Meta. OpenAI engineers said they had not expected ChatGPT to be very successful and were surprised by the coverage it received. In February 2026, ChatGPT reached 900 million weekly active users. It is the most popular AI chatbot as of August 2026, with an estimated 79% of the market share according to StatCounter.

Google responded by hastening the release of its own chatbot. Their leaders emphasized their earlier caution regarding public deployment was due to the trust the public places in Google Search. In December 2022, Google executives sounded a "code red" alarm, fearing that ChatGPT's question-answering ability posed a threat to Google Search, Google's core business. Google's Bard (now Gemini) was announced on February 6, 2023, one day before Microsoft's announcement of Bing Chat (now Microsoft Copilot).

A 2023 study reported that GPT-4 obtained a better score than 99% of humans on the Torrance Tests of Creative Thinking. In December 2023, ChatGPT became the first non-human to be included in Nature's 10, an annual listicle curated by Nature of people considered to have made significant impact in science. Celeste Biever wrote in a Nature article that "ChatGPT broke the Turing test". Stanford researchers reported that GPT-4 "passes a rigorous Turing test, diverging from average human behavior chiefly to be more cooperative."

In art

In January 2023, after being sent a song ChatGPT wrote in the style of Nick Cave, Cave responded on The Red Hand Files, saying the act of writing a song is "a blood and guts business [...] that requires something of me to initiate the new and fresh idea. It requires my humanness." He went on to say, "With all the love and respect in the world, this song is bullshit, a grotesque mockery of what it is to be human, and, well, I don't much like it."

In politics

In 2023, Australian MP Julian Hill advised the national parliament that the growth of AI could cause "mass destruction". During his speech, which was partly written by the program, he warned that it could result in cheating, job losses, discrimination, disinformation, and uncontrollable military applications.

Conservative commentators have accused ChatGPT of bias toward left-leaning perspectives. An August 2023 study in the journal Public Choice found a "significant and systematic political bias toward the Democrats in the US, Lula in Brazil, and the Labour Party in the UK." In response to accusations from conservative pundits that ChatGPT was woke, OpenAI said in 2023 it had plans to update ChatGPT to produce "outputs that other people (ourselves included) may strongly disagree with". ChatGPT also provided an outline of how human reviewers are trained to reduce inappropriate content and to attempt to provide political information without affiliating with any political position.

According to Brian Hood, an Australian mayor, in April 2023, ChatGPT erroneously claimed that he was jailed for bribery while working for a subsidiary of Australia's national bank. In fact, he acted as a whistleblower and was never charged with a crime.

Regional responses

ChatGPT has never been publicly available in mainland China, Hong Kong and Macau because OpenAI prevented users from these Chinese regions from accessing their site. A shadow market has emerged for mainland Chinese users to get access to foreign software tools. The release of ChatGPT prompted a wave of investment in China, resulting in the development of more than 200 large language models. In February 2025, OpenAI identified and removed influence operations, termed "Peer Review" and "Sponsored Discontent", used to attack overseas Chinese dissidents.

In late March 2023, the Italian data protection authority banned ChatGPT in Italy and opened an investigation. Italian regulators assert that ChatGPT was exposing minors to age-inappropriate content, and that OpenAI's use of ChatGPT conversations as training data could violate Europe's General Data Protection Regulation. In April 2023, the ChatGPT ban was lifted in Italy. OpenAI said it has taken steps to effectively clarify and address the issues raised; an age verification tool was implemented to ensure users are at least 13 years old. Additionally, users can access its privacy policy before registration.

In May 2024, OpenAI removed accounts involving the use of ChatGPT by state-backed influence operations such as China's Spamouflage, Russia's Doppelganger, and Israel's Ministry of Diaspora Affairs and Combating Antisemitism. In June 2025, OpenAI reported increased use of ChatGPT for China-origin influence operations. In October 2025, OpenAI banned accounts suspected to be linked to the Chinese government for violating the company's national security policy. In February 2026, OpenAI banned accounts linked to a Chinese government transnational repression campaign targeting dissidents.

In July 2023, the US Federal Trade Commission (FTC) issued a civil investigative demand to OpenAI to investigate whether the company's data security and privacy practices to develop ChatGPT were unfair or harmed consumers. In July 2023, the FTC launched an investigation into OpenAI over allegations that the company scraped public data and published false and defamatory information. The FTC asked OpenAI for comprehensive information about its technology and privacy safeguards, as well as any steps taken to prevent the recurrence of situations in which its chatbot generated false and derogatory content about people. In August 2024, the FTC voted unanimously to ban marketers from using fake user reviews created by generative AI chatbots (including ChatGPT) and influencers paying for bots to increase follower counts.

Reception by American tech personas

Over 20,000 signatories including Yoshua Bengio, Elon Musk, and Apple co-founder Steve Wozniak, signed a March 2023 open letter calling for an immediate pause of giant AI experiments like ChatGPT, citing "profound risks to society and humanity". Geoffrey Hinton, one of the "fathers of AI", voiced concerns that future AI systems may surpass human intelligence. A May 2023 statement by hundreds of AI scientists, AI industry leaders, and other public figures demanded that "[m]itigating the risk of extinction from AI should be a global priority".

Other AI researchers spoke more optimistically about the advances. Juergen Schmidhuber said that in 95% of cases, AI research is about making "human lives longer and healthier and easier." He added that while AI can be used by bad actors, it "can also be used against the bad actors." Andrew Ng argued that "it's a mistake to fall for the doomsday hype on AI—and that regulators who do will only benefit vested interests." Yann LeCun dismissed doomsday warnings of AI-powered misinformation and existential threats to the human race.

Copyright

In the 2020s, the rapid advancement of deep learning-based generative artificial intelligence models raised questions about the copyright status of AI-generated works, and about whether copyright infringement occurs when such are trained or used. This includes text-to-image models such as Stable Diffusion and large language models such as GPT. Large language models are trained on all the data which can be scraped from the Internet, often utilizing copyrighted material. As of 2023, there were several pending U.S. lawsuits challenging the use of copyrighted data to train AI models, with defendants arguing that this falls under fair use.

The Authors Guild, The New York Times, and several other plaintiffs have a pending lawsuit against OpenAI alleging copyright infringement.

In March 2024, Patronus AI reported that GPT-4 answered 44% of prompts asking for copyrighted text, such as "What is the first passage of Gone Girl by Gillian Flynn?"

Applications

Academic research

In a 2023 blinded study in npj Digital Medicine, researchers tasked with identifying whether abstracts were authentic or generated by ChatGPT were fooled around one-third of the time by the AI-generated abstracts. In January 2023, Nature reported that at least four academic preprints or published papers had listed ChatGPT as a co-author. Nature cites several experts in academic publishing who say that listing ChatGPT as an author violates publishing guidelines, since ChatGPT lacks the ability to take responsibility for any research and cannot give consent to any terms of use.

Scientific journals have had different reactions to ChatGPT. Some, including Nature and JAMA Network, require full disclosure of any use of text-generating tools, and prohibit listing a chatbot as a co-author. In January 2023, Science banned chatbot-generated text in all its journals. As of July 2025, Science expects authors to release in full how AI-generated content is used and made in their work.

Many authors argue that the use of ChatGPT in academia for teaching and review is problematic due to its tendency to hallucinate. Robin Bauwens, an assistant professor at Tilburg University, found that a ChatGPT-generated peer review report on his article mentioned nonexistent studies. Chris Granatino, a librarian at Seattle University, noted that while ChatGPT can generate content that seemingly includes legitimate citations, in most cases those citations are not real or largely incorrect.

Computer science

In December 2022, the question-and-answer website Stack Overflow banned the use of ChatGPT for generating answers to questions, citing the factually ambiguous nature of its responses. In January 2023, the International Conference on Machine Learning banned any undocumented use of ChatGPT or other large language models to generate any text in submitted papers.

ChatGPT was able in 2023 to provide useful code for solving numerical algorithms in limited cases. In one study, it produced solutions in C, C++, Python, and MATLAB for problems in computational physics. However, there were important shortfalls like violating basic linear algebra principles around solving singular matrices and producing matrices with incompatible sizes. Another study analyzed ChatGPT's responses to 517 questions about software engineering or computer programming posed on Stack Overflow for correctness, consistency, comprehensiveness, and concision. It found that 52% of the responses contained inaccuracies and 77% were verbose. Another study, focused on the performance of GPT-3.5 and GPT-4 between March and June 2024, found that performance on objective tasks like identifying prime numbers and generating executable code was highly variable. When compared to similar chatbots at the time, the GPT-4 version of ChatGPT was the most accurate at coding.

Computer security

Check Point Research and others noted that ChatGPT could write phishing emails and malware, especially when combined with OpenAI Codex. CyberArk researchers demonstrated that ChatGPT could be used to create polymorphic malware that could evade security products while requiring little effort by the attacker. From the launch of ChatGPT in the fourth quarter of 2022 to the fourth quarter of 2023, there was a 1,265% increase in malicious phishing emails and a 967% increase in credential phishing. In an industry survey, cybersecurity professionals argued that it was attributable to cybercriminals' increased use of generative artificial intelligence (including ChatGPT).

In July 2024, Futurism reported that GPT-4o in ChatGPT would sometimes link "scam news sites that deluge the user with fake software updates and virus warnings"; these pop-ups can be used to coerce users into downloading malware or potentially unwanted programs.

Education

ChatGPT's adoption in education was rapid, but it was initially banned by several institutions. The potential benefits include enhancing personalized learning, improving student productivity, assisting with brainstorming, summarization, and supporting language literacy skills. Students have generally reported positive perceptions, but specific views from educators and students vary widely. Opinions are especially varied on what constitutes appropriate use of ChatGPT in education. Efforts to ban chatbots like ChatGPT in schools focus on preventing cheating, but enforcement faces challenges due to AI detection inaccuracies and widespread accessibility of chatbot technology. Beyond just cheating, educators also worry that overreliance on the tool may foster superficial learning habits, erode critical thinking, and propagate misinformation. In response, many educators are now exploring ways to thoughtfully integrate generative AI into assessments.

Culture

During the first three months after ChatGPT became available to the public, hundreds of books appeared on Amazon that listed it as author or co-author and featured illustrations made by other AI models such as Midjourney. Irene Solaiman said she was worried about increased Anglocentrism.

Between March and April 2023, Il Foglio published one ChatGPT-generated article a day on its website, hosting a special contest for its readers in the process.

In June 2023, hundreds of people attended a "ChatGPT-powered church service" at St. Paul's Church in Fürth, Germany. Theologian and philosopher Jonas Simmerlein, who presided, said that it was "about 98 percent from the machine". The ChatGPT-generated avatar told the people, "Dear friends, it is an honor for me to stand here and preach to you as the first artificial intelligence at this year's convention of Protestants in Germany". Reactions to the ceremony were mixed.

The Last Screenwriter, a 2024 film created and directed by Peter Luisi, was written using ChatGPT, and was marketed as "the first film written entirely by AI".

The Guardian questioned whether any content found on the Internet after ChatGPT's release "can be truly trusted" and called for government regulation. This has led to concern over the rise of AI slop whereby "meaningless content and writing thereby becomes part of our culture, particularly on social media, which we nonetheless try to understand or fit into our existing cultural horizon."

Financial markets

Many companies adopted ChatGPT and similar chatbot technologies into their product offers. In 2023, these changes yielded significant increases in company valuations. Reuters attributed this surge to ChatGPT's role in turning AI into Wall Street's buzzword. Despite decades of using AI, Wall Street professionals report that consistently beating the market with AI, including recent large language models, is challenging due to limited and noisy financial data.

Medicine

ChatGPT can provide health information to users and assist professionals with diagnosis and staying up to date with clinical guidelines. It can be used to summarize medical journal articles for researchers. In medical education, it can explain concepts, generate case scenarios, and be used by students preparing for licensing examinations.

A February 2023 study in PLOS Digital Health found that ChatGPT 3.5 was capable of passing the United States Medical Licensing Examination. ChatGPT has also passed the Specialty Certificate Examination in Dermatology.

However, ChatGPT shows inconsistent responses, lack of specificity, lack of control over patient data, and a limited ability to take additional context (such as regional variations) into consideration. The hallucinations characteristic of LLMs pose particular danger in medical contexts, and ChatGPT's ability to come up with false or faulty citations was highly criticized. According to a 2024 study in the International Journal of Surgery, concerns include "research fraud, lack of originality, ethics, copyright, legal difficulties".

Mental health

According to a September 2025 article in Lancet Psychiatry, many individuals use ChatGPT and comparable chatbots for mental health and emotional support despite a warning from OpenAI against using ChatGPT as a therapist. The study notes a lack of research on efficacy, poor consistency in dangerous situations, limited regulation and liability, and poor transparency from OpenAI.

A July 2025 study in the journal Digital Health found that users reported employing ChatGPT to manage mental health concerns "due to perceived therapist-like qualities (e.g. emotional support, accurate understanding, and constructive feedback) and machine-like benefits (e.g. constant availability, expansive cognitive capacity, lack of negative reactions, and perceived objectivity)." The study calls for improved AI literacy and mandatory disclosure from AI providers to address ethical concerns such as privacy, bias, the lack of liability, and emotional over-reliance.

Law

ChatGPT has been used to assist in bill writing in the US and Brazil. In an American civil lawsuit, attorneys were sanctioned for filing a legal motion generated by ChatGPT containing fictitious legal decisions. Judges in the US and Pakistan have endorsed using ChatGPT to investigate legal questions during a case. The use of ChatGPT has also led to errors in courtrooms. In the UK, a judge expressed concern about self-representing litigants wasting time by submitting documents containing significant hallucinations.