A research firm, OpenAI has created a text generator model called GPT-2 which was considered “too dangerous” for public use.
The artificial intelligence system which some claim performs better than humans is able to generate realistic stories, poems and articles.
However, an updated version of the system has been released that can be used to create fake news or abusive spam on social media.
GPT-2 is trained on a data-set of eight million web pages, and is programmed to adjust to the style and content to the text that is provided to it.
AI Summit help boost Pakistani tech industry
"Due to our concerns about malicious applications of the technology, we are not releasing the trained model. As an experiment in responsible disclosure, we are instead releasing a much smaller model for researchers to experiment with," reported the firm.
The firm decided to release a version of the program that consists of lesser parameters in terms of phrases and sentences - than used during training.
Ma vs Musk: tech tycoons spar on future of AI
"I'm terrified of GPT-2 because it represents the kind of technology that evil humans are going to use to manipulate the population - and in my opinion that makes it more dangerous than any gun," according to BBC Article author Tristan Greene.
This month, OpenAI decided to grow the parameters, by offering a much broader database of training data.
This article was originally published on BBC.