Robots do many things to help humans, but due to efforts by Spanish researchers they now catch pedophiles in online chatrooms thanks to the Negobot.
The Negobot poses as a 14-year-old girl, and using artificial intelligence (AI) to realistically mimic the language of teenagers, strikes up conversations to catch pedophiles in online chatrooms.
Negobot, a “virtual Lolita” nicknamed for the eponymous main character of Vladimir Nabokov’s novel, starts off with a neutral emotion, but will adopt any of seven personalities according to the variety and intensity of interactions.
Negobot uses advanced decision-making strategies known as “game theory” in order to simulate realistic dialogues as they develop. Other features include taking the lead in conversations and remembering specific facts about what had been discussed previously and with whom.
Even more convincingly, this Negobot, a “conversational agent” uses child-like language and slang, along with spelling mistakes and contradictions to draw in the predator further.
Negobot would be used in chatrooms that suspected pedophiles are thought to be using, beginning new chats as a somewhat passive speaker. A behavioral adaptation follows in response to the methods used by the suspect to win over its trust and friendship.
Negobot can also appear offended or get more insistent if it’s not drawn into a conversation. It will also respond to more aggressive advances—like requests for personal information—by trying to find out more about the suspect, including phone numbers and social media accounts.
For example, if the suspect does not appear to be enticed into having a conversation, Negobot can appear offended or get more insistent.
Negobot will also respond to more aggressive advances - like requests for personal information - by trying to find out more about the suspect. This can include details such as their social network profile and mobile number, information which can then be used by police to start an investigation.
The project’s team at the University of Deusto says the software demonstrates a significant technological advance. According to Dr. Carlos Laorden, “chatbots” of the past tended to be quite predictable, with flat interest and behavior in a conversation.
Laorden said that was, “a problem when attempting to detect untrustworthy targets like pedophiles.”
John Carr, a UK government adviser on child protection, was grateful for measures that could help ease the workload of real-world policing. However, Carr cautioned that the software risked drawing people in to take actions they would not do otherwise.
“Undercover operations are extremely resource-intensive and delicate things to do. It’s absolutely vital that you don’t cross a line into entrapment which will foil any potential prosecution,” Carr said.
To date, the Negobot has been field-tested on Google’s chat service and could be translated into other languages, already attracting the attention of the Basque police force.
Researchers admit though that Negobot still has limitations and will need to be monitored to prevent significant error. While it is has broad conversational abilities, it is not yet sophisticated enough to decipher specific figurative features such as irony.