chiprook

← AI

AI52 mOctober 27, 2026, 20:00

Journalist used AI to fake a scientific paper and tested IEEE Access defenses

An IEEE Spectrum journalist used free ChatGPT and Gemini to generate a convincing paper on optical solitons without any real research, then tried to sneak it past IEEE Access's defenses. An estimated 57% of published academic articles showed signs of LLM influence by 2025, and over 12% of Nature Communications reviews were fully AI-generated.

IEEE Spectrum

Open source

Journalist used AI to fake a scientific paper and tested IEEE Access defenses

57% of papers in 2025 showed LLM influence, up from 12% in 2023
Over 12% of Nature Communications reviews in 2025 were fully AI-generated
Only 1 in 40 papers with AI traces formally disclosed its use
The fake paper was built with free ChatGPT and Gemini without manual edits

#OpenAI#Google#ChatGPT#Gemini

More stories in swipe mode🔥 Trending now: Anthropic files for IPO at a valuation of up to $2 trillion — more than most countries' GDP

Read next

AI

OpenAI exposes Bogus Bylines: 7 fake journalists, ~100 AI-written articles

AI

IEEE Xplore adds AI research assistant grounded in full-text papers

Policy

AI Firms Knew Chatbots Were an ‘Existential Threat’ to Journalists, Court Docs Show

AI

Redpine Opens 20 Million Research Papers to AI Agents via Licensing Deals