neutral
1 day agoMeta flags bias in major LLMs launched by OpenAI and Google, says Zuckerberg
Meta Platforms CEO Mark Zuckerberg has publicly stated that leading large-language models (LLMs) developed by OpenAI and Google LLC exhibit a “left-leaning” political bias.
Meta’s own model, Llama 4, is being updated to address such bias, with Zuckerberg emphasising that the issue affects the industry broadly, including models from OpenAI and Google. The announcement came amid growing regulatory and public scrutiny of generative-AI fairness.
Meta aims to set itself apart by promising more “balanced responses” and improved model transparency in its next release cycle.
neutral
1 day agoMeta flags bias in major LLMs launched by OpenAI and Google, says Zuckerberg
Meta Platforms CEO Mark Zuckerberg has publicly stated that leading large-language models (LLMs) developed by OpenAI and Google LLC exhibit a “left-leaning” political bias.
Meta’s own model, Llama 4, is being updated to address such bias, with Zuckerberg emphasising that the issue affects the industry broadly, including models from OpenAI and Google. The announcement came amid growing regulatory and public scrutiny of generative-AI fairness.
Meta aims to set itself apart by promising more “balanced responses” and improved model transparency in its next release cycle.
1 min read
82 words
Meta claims major AI models from OpenAI and Google show political bias and updates its Llama 4 model to respond more neutrally.
Meta Platforms CEO Mark Zuckerberg has publicly stated that leading large-language models (LLMs) developed by OpenAI and Google LLC exhibit a “left-leaning” political bias.
Meta’s own model, Llama 4, is being updated to address such bias, with Zuckerberg emphasising that the issue affects the industry broadly, including models from OpenAI and Google. The announcement came amid growing regulatory and public scrutiny of generative-AI fairness.
Meta aims to set itself apart by promising more “balanced responses” and improved model transparency in its next release cycle.
Meta Platforms CEO Mark Zuckerberg has publicly stated that leading large-language models (LLMs) developed by OpenAI and Google LLC exhibit a “left-leaning” political bias.
Meta’s own model, Llama 4, is being updated to address such bias, with Zuckerberg emphasising that the issue affects the industry broadly, including models from OpenAI and Google. The announcement came amid growing regulatory and public scrutiny of generative-AI fairness.
Meta aims to set itself apart by promising more “balanced responses” and improved model transparency in its next release cycle.
Companies:
Meta Platforms
OpenAI
Google LLC
Tags:
ai
global
ai
global
meta
openai
bias
Nov 8, 2025 • 20:08 IST