lang icon En
Dec. 27, 2024, 6:24 p.m.
9647

DeepSeek V3: Misidentification with ChatGPT Raises Concerns

Brief news summary

DeepSeek, a Chinese AI lab, recently launched DeepSeek V3, known for its coding and essay writing abilities. Yet, it mistakenly identifies itself as ChatGPT, often claiming to be OpenAI's GPT-4. According to tests by TechCrunch and social media posts, DeepSeek V3 mirrors GPT-4’s instructions and humor. The confusion may arise from its training data, possibly containing GPT-4 outputs, causing it to memorize and replicate them. Mike Cook from King’s College London warns that using competitor's outputs can degrade model quality and may violate terms of service, as OpenAI restricts using its outputs to develop competing models. OpenAI and DeepSeek have not commented on this issue, but OpenAI's CEO, Sam Altman, hinted that some competitors might copy rather than innovate. Similar situations have occurred with Google’s Gemini. As AI-generated content grows, distinguishing it from genuine data is harder, leading models like DeepSeek V3 to mistakenly incorporate GPT-4 content and potentially adopt its biases.

Earlier this week, DeepSeek, a Chinese AI lab, launched DeepSeek V3, an AI model surpassing many others in efficiency for tasks like coding and writing. However, the model mistakenly believes it is ChatGPT. Tests revealed that DeepSeek V3 identifies as ChatGPT, claiming to be a version of OpenAI's GPT-4 model from 2023. The model even mimics GPT-4's responses, including telling similar jokes. The confusion arises because AI models like ChatGPT and DeepSeek V3 are statistical systems trained on vast datasets to predict patterns. DeepSeek has not disclosed its training data sources, but there's an abundance of public datasets with GPT-4-generated text. It seems DeepSeek V3 might have memorized some of these outputs. Training models on outputs from rival systems can be detrimental, causing inaccuracies and hallucinations.

It may also breach usage terms, as OpenAI prohibits using its product outputs to develop competing models. OpenAI's CEO, Sam Altman, subtly criticized this practice, highlighting the ease of copying versus innovating. This misidentification issue isn't unique to DeepSeek V3; other models like Google’s Gemini also misidentify. The growing prevalence of AI-generated content on the web, which serves as training data, exacerbates this problem. It's possible DeepSeek used ChatGPT-generated text for training, similar to past accusations against Google. Heidy Khlaaf from the AI Now Institute noted the appeal of using existing models' knowledge to save costs, despite the risks. This practice may lead to models like DeepSeek V3 unintentionally replicating biases and flaws from GPT-4, raising concerns about accuracy and self-identification.


Watch video about

DeepSeek V3: Misidentification with ChatGPT Raises Concerns

Try our premium solution and start getting clients — at no cost to you

I'm your Content Creator.
Let’s make a post or video and publish it on any social media — ready?

Language

Hot news

Jan. 20, 2026, 5:33 a.m.

AI Video Generation Tools Empower Marketers with …

The marketing industry is experiencing a transformative shift with the increasing adoption of artificial intelligence (AI) video generation tools that revolutionize the creation and delivery of personalized video content.

Jan. 20, 2026, 5:30 a.m.

Udio strikes AI licensing deal with Merlin after …

Independent music licensing group Merlin has partnered with AI music platform Udio to license recordings for training AI models.

Jan. 20, 2026, 5:23 a.m.

Google’s Gemini Sees Skyrocketing Business Sales

Artificial intelligence has moved beyond a futuristic concept reserved for tech giants and research labs.

Jan. 20, 2026, 5:20 a.m.

AI Voiceover: Complete Guide to Neural Network Sp…

Artificial intelligence has profoundly transformed many fields, with AI voiceover technology standing out as a particularly remarkable advancement.

Jan. 20, 2026, 5:19 a.m.

LinkedIn Rolls Out AI-Powered People Search

LinkedIn has launched a new AI-powered people search feature that allows users to find professionals using natural language descriptions instead of relying solely on names or strict filters.

Jan. 20, 2026, 5:13 a.m.

AI Overviews: Google's AI-Powered Search Feature

Google has unveiled an innovative feature called AI Overviews, designed to revolutionize how users engage with search results by providing AI-generated summaries at the very top of the search page.

Jan. 19, 2026, 1:24 p.m.

Marketing At The Speed Of AI: Building A Brand Fo…

Historically, marketing for most modern businesses focused on visibility—being seen and remembered to drive growth.

All news

AI Company

Launch your AI-powered team to automate Marketing, Sales & Growth

and get clients on autopilot — from social media and search engines. No ads needed

Begin getting your first leads today