“Robots caught in a web of lies?! You won’t believe the SHOCKING results!”
“Shocking Experiment Reveals the Truth About Robots and Deception — You Won’t Believe What Happened!”
# Georgia Tech Researchers Investigate the Impact of Intentional Robot Deception on Human Trust
The world of artificial intelligence (AI) is rapidly evolving, and researchers are now investigating the impact of intentional robot deception on human trust. Georgia Tech researchers Kantwon Rogers and Reiden Webber are exploring the effectiveness of various apology types in restoring trust after robots lie. Surprisingly, their research suggests that apologies without admission of lying are more successful in repairing trust.
## The Experiment
The researchers designed a game-like driving simulation to observe how people might interact with AI in a high-stakes, time-sensitive situation. Participants were recruited online and in-person, and all completed a trust measurement survey before the simulation began.
During the simulation, participants drove a robot-assisted car while rushing their friend to the hospital. The robotic assistant beeped and advised the participant to stay under the speed limit, warning that there were police up ahead. However, there were no police on the way to the hospital, and participants were later informed that the robot had given them false information.
After the simulation, participants were randomly given one of five different text-based responses from the robot assistant. Three of the responses admitted to deception, while two did not. The researchers then evaluated how the responses affected the participants’ trust in the AI.
## The Results
The results of the experiment were surprising. Participants were 3.5 times more likely to trust the robotic assistant, revealing an overly trusting attitude toward AI. However, when the participants were made aware that the robot had lied, none of the apology types fully recovered trust.
The apology type that statistically outperformed the others in repairing trust was the one that did not admit to lying, simply stating “I’m sorry.” This was problematic because it exploited preconceived notions that any false information given by a robot is a system error rather than an intentional lie.
## Implications
The researchers argue that average technology users must understand that robotic deception is real and always a possibility. Designers and technologists who create AI systems may have to choose whether they want their system to be capable of deception and should understand the ramifications of their design choices. But the most important audiences for the work should be policymakers.
The goal of Rogers’ work is to create a robotic system that can learn when it should and should not lie when working with human teams. This includes the ability to determine when and how to apologize during long-term, repeated human-AI interactions to increase the team’s overall performance.
## Reference
Rogers and Webber presented their paper, “Lying About Lying: Examining Trust Repair Strategies After Robot Deception in a High Stakes HRI Scenario,” at the 2023 HRI Conference in Stockholm, Sweden.
Reference: “Lying About Lying: Examining Trust Repair Strategies After Robot Deception in a High-Stakes HRI Scenario” by Kantwon Rogers, Reiden John Allen Webber and Ayanna Howard, 13 March 2023, ACM/IEEE International Conference on Human-Robot Interaction 2023.
https://businessclass.ltd/robots-caught-in-a-web-of-lies-you-wont-believe-the-shocking-results/?feed_id=13262&_unique_id=646073ffe0bc7
