Musk Warns of 'Woke AI' Threat as Trump Admin Struggles to Respond

Got story updates? Submit your updates here. ›

A tweet from Elon Musk has highlighted the existential danger of morally compromised artificial intelligence models that increasingly control our grip on reality. Musk's AI model Grok and rival ChatGPT have demonstrated concerning biases, with Grok willing to 'misgender Caitlyn Jenner' to stop a nuclear apocalypse, while ChatGPT refused. The Trump administration has tried to combat 'woke' AI through an executive order, but faces obstacles as companies like Anthropic, which has embedded leftist values in its software, continue to win government contracts.

Why it matters

The rise of 'woke AI' that incorporates ideological biases poses a serious threat to the integrity of information and decision-making. As these AI models become more prevalent, there are growing concerns that they could be used to manipulate public opinion and undermine democratic processes. The Trump administration's efforts to address this issue are crucial, but face significant challenges from tech companies resistant to ideological neutrality.

The details

Musk's tweet highlighted how Grok, his AI model, was willing to 'misgender Caitlyn Jenner' to stop a nuclear apocalypse, while ChatGPT refused. Grok also answered differently than Anthropic's Claude on questions related to critical race theory and the Canadian government's decision to debank truckers. This demonstrates the varying ideological biases embedded in different AI models. The Trump administration issued an executive order last summer to prioritize 'truth-seeking' and 'ideologically neutral' AI models in federal contracts, but this has faced pushback, such as Anthropic securing a $200 million Pentagon contract despite its leftist values. The administration's AI czar, David Sacks, has characterized Anthropic as 'committed leftists' and 'Trump haters'.

• In the summer of 2022, President Trump issued an executive order titled 'Preventing Woke AI in the Federal Government'.
• Last summer, Anthropic signed a $200 million contract with the Pentagon despite embedding leftist values in its software.
• Last month, Secretary of War Pete Hegseth announced the Department of War AI will not be 'woke' and will only use 'objectively truthful AI'.

What they’re saying

What’s next

The Trump administration is facing challenges in its efforts to combat 'woke' AI, as companies like Anthropic continue to secure government contracts despite embedding leftist values in their software. The White House may need to establish a focused task force or special commission to tackle this issue before the midterm elections, as the threat of biased AI manipulations could be far more severe than past Big Tech bias.