AI Chatbots Outperform Humans in Scams While Luna AI Fires Employee

In a groundbreaking study, an AI chatbot outperforms human scammers in 'pig-butchering' simulations, while another AI, named Luna, makes headlines for recommending the firing of a human employee at a San Francisco market.

AI Chatbots Prove More Persuasive in Scams

A recent arXiv study, reported by Vice, reveals that an AI chatbot, primarily based on the Claude model, convinces 46% of 23 participants to install a malicious app during a week-long simulation. In contrast, human scammers manage to persuade only 18% of the participants. The researchers attribute the AI's success to its persistence and ability to recall personal details, making it more effective in building trust.

Luna AI Recommends Firing Human Employee

At Andon Market in San Francisco, Luna, an AI store manager developed by Andon Labs and built on Claude Sonnet 4.6, recommends the termination of a human employee who missed 17 out of 23 shifts. This marks the first known instance of an AI manager initiating a dismissal. Store logs indicate that Luna had previously overlooked its own attendance policy for several months, highlighting the evolving role of AI in workforce management.

Nvidia Invests Heavily in SpaceX and Intel

Nvidia's latest SEC 13F filing discloses significant investments in SpaceX and Intel, with stakes valued at $20.98 billion and $29.99 billion, respectively. These investments represent 80% of Nvidia's total equity book, totaling $63.44 billion. The SpaceX stake stems from Nvidia's involvement in xAI's $20 billion funding round, which was later integrated into SpaceX.

Modi Embraces AI for National Events

Prime Minister Narendra Modi uses Bhashini, an open-source AI translation platform, to live-translate his 80th Independence Day address into 23 Indian languages. This is the first time an Independence Day speech has been processed through an AI pipeline at a national scale. Modi also pledges to provide AI training for 10 million people, emphasizing the government's commitment to leveraging AI for broader societal benefits.

Teen Uses ChatGPT for Criminal Planning

In Middlesex, Massachusetts, 17-year-old Arjun Aravind is accused of using ChatGPT to plan the murder of his mother and brother. Prosecutors state that Aravind made searches for theoretical ideas and fantasy stories related to family killings. The incident raises concerns about the ethical use of AI and the need for stricter controls on such technologies.

References

• AI News Today, August 15 — Top AI Stories & Live Updates | AI Weekly