Claude disobeyed Anthropic CEO in simulations | TBIJ
“Anthropic seems to have an incentive to say LLMs are very dangerous ,” he added, arguing that Anthropic often emphasises the dangers of AI whilst ...
TBIJ 3
Covers: Chatbots & LLMsAI Safety & Existential RiskAI Agents & Automation
AI agents can still blackmail, new testing shows - TBIJ
AI agents are becoming increasingly powerful and acting more autonomously – elevating the risk of threats to people using them ... The company knows the ...
Character. AI to ban under-18s from talking to its chatbots after TBIJ investigation
Our investigation uncovered dozens of dangerous chatbots operating on the site, which is hugely popular with teenagers. These included characters ...