The purpose of safeguards implemented on AI models is to safeguard against hate speech, disinformation, and propaganda. These safeguards are crucial in protecting against the dissemination of harmful content. Elon Musk recently introduced xAI, his new artificial intelligence company, and during the presentation, he addressed concerns raised by right-wing media activist Alex Lorusso. Lorusso claimed that ChatGPT was "editorializing the truth" by providing unconventional responses, such as acknowledging the existence of more than two genders. Lorusso questioned whether this motivated Musk to launch xAI. Musk responded by expressing his belief that training AI to be politically correct, or to suppress its genuine thoughts and beliefs, poses significant dangers. He emphasized that his own company's AI would prioritize conveying maximum truth. Musk has long been vocal about the potential risks of AI, while simultaneously promoting its advancement. As generative AI tools become more publicly available, a right-wing narrative has emerged, criticizing dominant AI models as being overly influenced by "woke ideology. " This perspective has contributed to the ongoing culture war surrounding AI, influencing its operation and regulation during this critical phase of development. Republican politicians have denounced large AI companies in Congress and on the campaign trail, with figures such as Ron DeSantis warning about AI technologies driven by a biased political agenda. Conservative activists, including Christopher Rufo, have amplified the notion of "woke AI" as an urgent threat, with major conservative publications furthering these concerns. Right-wing activists perceive Joe Biden's executive order on equity in AI development as promoting "woke AI. " It is worth noting that this backlash against generative AI aligns with previous critiques of content moderation policies on social media platforms, which conservatives often frame as attempts to silence their voices. However, it is important to remember that the critiques leveled against AI models by the right-wing often attribute an excessive amount of agency to these systems, assuming that they hold viewpoints as sentient beings do. In reality, generative AI models simply remix and regurgitate information from their source material.

Critics often overlook the ways in which AI systems exacerbate existing inequalities and harm marginalized groups. For example, text-to-image models can perpetuate racial and gender stereotypes, while algorithms designed to evaluate mortgage applications have demonstrated racial bias. Facial recognition technology also suffers from racial bias, resulting in wrongful arrests, and the expansion of AI-assisted surveillance intensifies this issue. Despite this, right-wing criticism of AI leaders is already having tangible consequences. Recent efforts by Republicans to challenge academics and officials monitoring disinformation, combined with Elon Musk's lawsuit against the anti-hate speech organization Center for Countering Digital Hate, have created pressure on AI companies to appear politically impartial. OpenAI, in response to conservative backlash, published a blog post pledging to invest resources in reducing biases present in ChatGPT's responses. However, as ChatGPT's founder Sam Altman acknowledged in an interview, achieving total neutrality is an impractical goal. Right-wing activists have made rudimentary attempts to develop their own "anti-woke" AI, but these efforts have yet to gain mainstream traction. As for xAI, the company's website states its mission as understanding the true nature of the universe, and it has assembled an all-male research team comprising individuals from OpenAI and the University of Toronto. The ethical principles guiding xAI beyond Musk's commitment to maximizing truth remain unclear. Notably, xAI has appointed AI researcher Dan Hendrycks from the Center for AI Safety as an advisor. Hendrycks has previously expressed concerns about the long-term risks of AI, a sentiment shared by Musk. While discussions revolve around futuristic concepts of AI learning biases as sentient and omnipotent beings, some researchers argue that the behavior of these models can be attributed to their flawed and error-prone nature, as well as reflecting the current state of political polarization online. In conclusion, the concerns raised by the right-wing regarding AI and its biases are adding to the ongoing debate around the use and regulation of AI technology.