Arvind Narayanan, a computer scientist from Princeton, delivered a keynote address at the Computation + Journalism 2026 Symposium. The event was held at Northwestern University on October 2-3, 2026, under the theme AI and Journalism: Skating Where the Puck is Going.
Narayanan stated that AI tools could evolve into truth oracles within five to ten years. He advised newsrooms to shift their focus from current AI limitations to the capabilities of models expected in the coming decade.
Current AI models are often identified by readers due to factual errors and recognizable writing styles. Narayanan believes these weaknesses are temporary and will diminish as technology advances, making AI output appear more credible.
A 2026 survey by the Pew Research Center, involving 5,119 US adults, indicated that 10% of Americans view chatbots as authoritative sources of truth. Additionally, 29% of respondents identified chatbots as their primary association with AI.
The symposium addressed challenges facing the media industry, including the proliferation of low-quality machine-generated content. Narayanan suggested that journalists must increase their technical expertise to effectively oversee AI-driven information systems.
