A wordsmith AI could be a fake news factory

Feb 18, 2019, 3:26 AM EST
(Source: Rebecca Siegel/flickr)
(Source: Rebecca Siegel/flickr)

Algorithms are mastering the art of lying, and this could wreak havoc. This isn’t an exaggeration but a forewarning of what might be in the store if AI programs, adept at writing prose, get working in the real world.

A team at OpenAI, a research institute based in San Francisco, has developed an AI system, which was supposed to be a general-purpose language program, but has turned out to be a little too skilled with its flair for generating “generate malicious-esque content” on cue, writes MIT Technology Review.

Jack Clark, policy director at OpenAI, admitted that its AI model, called GPT2, could automate fake news generation and dissemination on the internet and social media, giving new dimensions to a challenge that already threatens elections and democratic functioning across the world, notes The Guardian.

For the fear of misuse of this “deepfakes for text,” the creators have decided to withhold the research details from public. Nevertheless, the fact that AI can imitate the literary skills of humans is a disturbing proposition for those whose bread and butter is tied to words.