Study finds AI chatbots are better at scamming people than humans
KeanuRave100 · reddit · 2026-08-16
A study reported by Vice reveals that AI-driven chatbots are more successful at phishing scams than human scammers. The research indicates that LLMs can send targeted scam messages at higher frequency and scale, and are more efficient at inducing victims to click malicious links or divulge sensitive information. This finding has sparked new concerns regarding AI misuse and regulation.
Related event: Study: AI Chatbots Outperform Humans in Scams(2 posts)→
More from Safety
- Dario Amodei: AI Could Cure Most Diseases in 5-10 Years, But Needs Regulation — rohanpaul_ai · 2026-08-17
- Commentary: OpenAI and Hugging Face incident was not a PR stunt — OwariDa · 2026-08-17
- Man used Grok to create 7,000 fake explicit images of stepdaughter, raising AI safety concerns — SatelliteNetSec · 2026-08-17
- Expert Concerns Embedded Devices as a Weak Link in AI Security — johnowhitaker · 2026-08-17
- Secure enterprise agent credentials with Vault integration — blaizedsouza · 2026-08-17
- Zvi Proposes Generalizing AI Constitution Definitions — TheZvi · 2026-08-17