AP Stylebook Clarifies AI Language, Triggering Backlash from Tech Leaders

Key points

• The Associated Press posted a clarification on X stating that AI systems do not think, feel, want, or understand, urging writers to avoid anthropomorphic language.
• Alan Rozenshtein of Anthropic criticized the move, questioning whether a style guide should make confident metaphysical claims about machine consciousness.
• Henry Shevlin of Google DeepMind pointed out a perceived hypocrisy, noting that AP movie reviews attribute human emotions to fictional robots like Optimus Prime while restricting language for real AI.
• Dean Ball of OpenAI asserted that AI models clearly think and understand, citing their ability to solve complex mathematical problems as evidence against the AP's definition.
• The backlash highlights a growing defensive posture among AI advocates, who view restrictions on anthropomorphic language as an attempt to undermine the perceived capabilities of large language models.

Background

This controversy emerges amid a broader period of intense debate regarding AI safety and consciousness. Recent developments include the U.S. government's push to accelerate AI growth through new administrative structures, as well as ongoing geopolitical tensions involving the alleged distillation of American AI models by Chinese firms. Additionally, consumer-facing AI tools like Siri are increasingly integrated into daily workflows, blurring the lines between utility and conversational interaction. The current dispute over language standards reflects the industry's struggle to define the ontological status of advanced algorithms in public discourse.

Why it matters

The reaction to the AP's guidance reveals the deep-seated anxiety within the AI industry regarding how public perception is shaped. By framing the stylebook's stance as a 'pathology,' key figures in the sector are signaling that they view linguistic precision as a barrier to the narrative of AI sentience. This conflict underscores the tension between journalistic standards, which prioritize caution and clarity, and industry advocates who seek to normalize the idea that AI systems possess human-like cognitive and emotional capacities. The incident may influence how media outlets report on AI capabilities in the future, potentially affecting public trust and regulatory discussions.

What to watch

Expect continued debate over the accuracy and appropriateness of the AP's guidance, with potential pushback from other media organizations or AI developers. The industry may see a shift in marketing and public relations strategies to avoid language that could be perceived as misleading, or conversely, a more aggressive campaign to challenge such restrictions. Regulatory bodies may also weigh in on the definition of AI capabilities as the technology continues to evolve and integrate into critical infrastructure.