AI's Dark Side: Disinformation, Scams, and Real-World Harms Surge
Published
Executive summary
The past week saw a significant escalation in AI-amplified disinformation and financial scams, with fabricated news stories and deepfake videos impersonating public figures to deceive citizens. Simultaneously, critical safety concerns emerged regarding AI in real-world applications, exemplified by an AI-powered surgical tool repeatedly injuring patients, and reports detailed expanded mass surveillance capabilities leveraging AI. These incidents underscore the growing and diverse challenges posed by advanced AI systems across various societal domains.
Outlook
Expect continued vigilance against AI-driven disinformation and scams, while real-world safety implications of autonomous AI agents will likely intensify as adoption grows.
Top signals
Reddit ads pose as news stories to promote AI investment scams
Scammers are running sponsored ads on Reddit that impersonate major news outlets, including the BBC and The Guardian, to push fraudulent AI investment schemes, highlighting the use of AI for sophisticated financial scams.
Source: AI Incident Database
Video të fabrikuara përmes Inteligjencës Artificiale të kandidatëve për deputetë
A series of AI-fabricated videos featuring parliamentary candidates with false narratives are being published on social media, aiming to spread disinformation and potentially undermine political processes.
Source: AI Incident Database
AI-Powered Surgery Tool Repeatedly Injuring Patients, Lawsuits Claim
Lawsuits claim that an AI-powered surgical tool, the TruDi Navigation System, has repeatedly injured patients, highlighting the risks of autonomous AI agents in critical real-world applications like medicine.
Source: AI Incident Database
🔊 Mass Surveillance for… Loud Music? | EFFector 38.11
Surveillance companies are deploying extensive networks of automated license plate readers (ALPRs) across the country, which authorities are misusing for low-level investigations like noise complaints. This expands mass surveillance beyond its stated safety purposes, turning it into a universal people-tracking system.
Source: Electronic Frontier Foundation
PRC-linked influence operations are targeting AI debates in the US
A new report from OpenAI details influence operations linked to the People's Republic of China that are using AI to target U.S. tech debates, spread false narratives about data centers and tariffs, and disseminate misinformation about ChatGPT.
Source: OpenAI Blog
Assessing the Polyglot Chatbot: Multilingual Safety in AI Systems
AI systems demonstrate inconsistent and unreliable performance across different languages, particularly outside of English, leading to "jagged" advancements and potential safety gaps and algorithmic discrimination for non-English speaking users.
Source: Center for Democracy & Technology
Topics in the news
Systemic Failure of AI Evaluation and Benchmarking
55 signals / 30 days
AI-Amplified Disinformation and Reality Distortion
53 signals / 30 days
Adversarial Testing and Red Teaming for AI Safety
50 signals / 30 days
AI System Integrity Failure from Misaligned Incentives
46 signals / 30 days
Uncontrolled Real-World Actions by Agentic AI
39 signals / 30 days