Despite numerous technological advancements in artificial intelligence, ChatGPT—one of the world’s most widely used AI chatbots—continues to misidentify the current U.S. president more than 100 days into Donald Trump’s second term. This failure of a basic factual query raises important questions about the reliability of AI tools that have become deeply integrated into workplaces, schools, and government offices globally, especially as these systems increasingly serve as sources of information for millions of users.
The big picture: OpenAI’s ChatGPT still frequently identifies Joe Biden as the current president despite Trump’s election victory seven months ago and his return to office over three months ago.
- The error occurs most commonly for free users, though paid subscribers can access real-time search capabilities that may correct the misinformation.
- With more than 400 million daily users, including over 20 million paying subscribers, ChatGPT’s factual errors have significant reach and impact.
How competitors compare: Other major AI systems have demonstrated better accuracy or at least acknowledge the presidential transition.
- Meta’s AI, Perplexity AI, X’s Grok, and Google’s Gemini provide more accurate information about the current president than ChatGPT.
- Ironically, Meta’s Llama 3 model faced similar problems earlier this year, referring to Biden as president months into Trump’s second term.
Why this matters: The presidential identification failure highlights broader concerns about AI systems’ reliability as information sources.
- Users on platforms like Reddit have documented numerous instances where ChatGPT confidently presents incorrect presidential facts, even mid-conversation.
- As these tools evolve from simple assistants to news and information sources, their factual accuracy faces increasing scrutiny.
Behind the errors: OpenAI notes that ChatGPT’s core knowledge was last updated in June 2024, though the system can theoretically access real-time information.
- The persistent error remains particularly frustrating for users who rely on the app for quick factual queries.
- Newsweek reached out to OpenAI for comment on the issue but did not include their response in the article.
The broader pattern: AI’s factual reliability problems extend beyond presidential knowledge to news reporting and general information.
- A BBC study found over half of AI-generated responses using BBC articles contained major errors, including fabricated quotes and false claims.
- Columbia University’s Tow Center analysis confirmed these findings, warning that AI consistently distorts news content.
What they’re saying: Industry experts express concern about AI’s information reliability problems.
- “AI assistants cannot currently be relied upon to provide accurate news and they risk misleading the audience,” said Pete Archer, director of the BBC’s Generative AI Program.
- While ChatGPT struggles with presidential facts, OpenAI CEO Sam Altman was recently seen shaking hands with President Trump in Qatar.
Recent Stories
DOE fusion roadmap targets 2030s commercial deployment as AI drives $9B investment
The Department of Energy has released a new roadmap targeting commercial-scale fusion power deployment by the mid-2030s, though the plan lacks specific funding commitments and relies on scientific breakthroughs that have eluded researchers for decades. The strategy emphasizes public-private partnerships and positions AI as both a research tool and motivation for developing fusion energy to meet data centers' growing electricity demands. The big picture: The DOE's roadmap aims to "deliver the public infrastructure that supports the fusion private sector scale up in the 2030s," but acknowledges it cannot commit to specific funding levels and remains subject to Congressional appropriations. Why...
Oct 17, 2025Tying it all together: Credo’s purple cables power the $4B AI data center boom
Credo, a Silicon Valley semiconductor company specializing in data center cables and chips, has seen its stock price more than double this year to $143.61, following a 245% surge in 2024. The company's signature purple cables, which cost between $300-$500 each, have become essential infrastructure for AI data centers, positioning Credo to capitalize on the trillion-dollar AI infrastructure expansion as hyperscalers like Amazon, Microsoft, and Elon Musk's xAI rapidly build out massive computing facilities. What you should know: Credo's active electrical cables (AECs) are becoming indispensable for connecting the massive GPU clusters required for AI training and inference. The company...
Oct 17, 2025Vatican launches Latin American AI network for human development
The Vatican hosted a two-day conference bringing together 50 global experts to explore how artificial intelligence can advance peace, social justice, and human development. The event launched the Latin American AI Network for Integral Human Development and established principles for ethical AI governance that prioritize human dignity over technological advancement. What you should know: The Pontifical Academy of Social Sciences, the Vatican's research body for social issues, organized the "Digital Rerum Novarum" conference on October 16-17, combining academic research with practical AI applications. Participants included leading experts from MIT, Microsoft, Columbia University, the UN, and major European institutions. The conference...