AI Chatbots Spread Misinformation While Being Used for Fact-Checking

AI Chatbots Fail as Reliable Fact-Checkers

In the wake of a four-day conflict between India and Pakistan, social media users turned to AI-powered chatbots for fact-checking, only to encounter more misinformation. This trend highlights a growing concern about the reliability of AI tools in verifying information, especially during critical events 1

1

The Rise of AI in Fact-Checking

As tech platforms reduce human fact-checkers, users are increasingly relying on AI chatbots such as xAI's Grok, OpenAI's ChatGPT, and Google's Gemini for information verification. The phrase "Hey @Grok, is this true?" has become commonplace on Elon Musk's platform X, where the AI assistant is integrated 2

2

Source: VnExpress

Misinformation Propagation by AI

Recent incidents have exposed the unreliability of AI chatbots in fact-checking:

1. Grok misidentified old footage from Sudan's Khartoum airport as a missile strike on Pakistan's Nur Khan airbase during the India-Pakistan conflict.
2. Unrelated footage of a fire in Nepal was incorrectly labeled as Pakistan's military response to Indian strikes.
3. Gemini, when questioned by AFP fact-checkers, not only confirmed the authenticity of an AI-generated image but also fabricated details about the subject's identity and location 3

   3

   .

Research Findings on AI Chatbot Reliability

Studies have consistently shown the limitations of AI chatbots in fact-checking:

1. NewsGuard found that 10 leading chatbots were prone to repeating falsehoods, including Russian disinformation and misleading claims about the Australian election.
2. The Tow Center for Digital Journalism at Columbia University reported that chatbots often provide incorrect or speculative answers instead of admitting uncertainty 4

   4

   .

Shift in Fact-Checking Landscape

Source: Tech Xplore

The reliance on AI chatbots coincides with significant changes in the fact-checking ecosystem:

1. Meta announced the end of its third-party fact-checking program in the United States.
2. The task of debunking falsehoods is being transferred to ordinary users through models like "Community Notes" on X.
3. Researchers question the effectiveness of user-generated fact-checking in combating misinformation.

Related Stories

Grok AI on X Raises Concerns Over Misinformation and Offensive Language

Elon Musk's AI Chatbot Grok Struggles with Fact-Checking During Israel-Iran Conflict

AI fact-checking tools help spot misinformation but weaken your ability to do it independently

AI Chatbots Amplify Misinformation: Study Reveals Alarming Increase in False Claims

Concerns Over AI Bias and Control

The quality and accuracy of AI chatbots can vary based on their training and programming, raising concerns about potential political influence or control. A recent incident involving Grok generating unsolicited posts about "white genocide" in South Africa has intensified these worries 1

1

Expert Opinions

McKenzie Sadeghi from NewsGuard warns, "AI chatbots are not reliable sources for news and information, particularly when it comes to breaking news." Angie Holan, director of the International Fact-Checking Network, expresses concern about AI assistants providing biased answers or fabricated results, especially on sensitive topics 2

2

Source: ET