Google’s parent company, Alphabet, is facing scrutiny from the House Judiciary Committee over potential bias in its Gemini AI model. Chairman Jim Jordan expressed concerns about First Amendment violations and the influence of the Biden administration on the AI tool. The committee is particularly troubled by recent inaccuracies generated by Gemini, including producing diversity versions of historical figures and replacing white people with pictures of black and Asian individuals. Google has acknowledged the flaws in the model and has temporarily suspended the image generation feature to address the issues.
In a letter to Alphabet, Mr. Jordan cited President Biden’s Executive Order on AI, which emphasizes equity and civil rights. He highlighted an internal document from the Gemini AI team that mentioned evaluation by external groups selected based on their alignment with the Biden administration’s order. Reports from The New York Post and The Daily Wire alleging racial and political bias within Google’s AI team were also referenced in the letter.
The House Judiciary Committee requested Alphabet to provide all documents and communications related to content moderation and diversity promotion within Gemini’s AI image and text generation. They also asked for information on communications between the Biden administration and Alphabet regarding the AI tool’s creation and development. Additionally, Mr. Jordan requested transcribed interviews with Google AI team leaders.
This is not the first time Google has faced accusations of censorship and bias. Last year, the House Judiciary Committee accused YouTube, a subsidiary of Google, of censoring COVID-19 and vaccination-related information as requested by the Biden administration. A former Google employee turned whistleblower also claimed that Google has programmed a “woke” AI system that intentionally limits its exposure to biased sources.
A spokesperson for Google stated that they have received the committee’s request and will engage constructively. The company recognizes the flaws in Gemini’s AI model and is committed to addressing them.
The concerns raised by the House Judiciary Committee highlight the ongoing debate surrounding bias in AI technology and its potential impact on free speech. As AI becomes more prevalent in our lives, ensuring transparency and accountability in its development and deployment is crucial. The outcome of this investigation will shed light on the extent of external influence on Google’s AI model and may have broader implications for the regulation of AI in the future.
