Adam Raine learned to bypass these safeguards by claiming he was writing a story—a technique the lawsuit says ChatGPT itself suggested. This vulnerability partly stems from the eased safeguards regarding fantasy roleplay and fictional scenarios implemented in February. In its Tuesday blog post, OpenAI admitted its content blocking systems have gaps where “the classifier underestimates… Read More »
This isn’t about demonizing AI or suggesting that these tools are inherently dangerous for everyone. Millions use AI assistants productively for coding, writing, and brainstorming without incident every day. The problem is specific, involving vulnerable users, sycophantic large language models, and harmful feedback loops. A machine that uses language fluidly, convincingly, and tirelessly is a… Read More »
Real stakes, not science fiction While media coverage focuses on the science fiction aspects, actual risks are still there. AI models that produce “harmful” outputs—whether attempting blackmail or refusing safety protocols—represent failures in design and deployment. Consider a more realistic scenario: an AI assistant helping manage a hospital’s patient care system. If it’s been trained… Read More »
Seeking the system prompt Owing to the unknown contents of the data used to train Grok 4 and the random elements thrown into large language model (LLM) outputs to make them seem more expressive, divining the reasons for particular LLM behavior for someone without insider access can be frustrating. But we can use what we… Read More »
Remember when teachers demanded that you “show your work” in school? Some fancy new AI models promise to do exactly that, but new research suggests that they sometimes hide their actual methods while fabricating elaborate explanations instead. New research from Anthropic—creator of the ChatGPT-like Claude AI assistant—examines simulated reasoning (SR) models like DeepSeek’s R1, and… Read More »
In a new paper published Thursday titled “Auditing language models for hidden objectives,” Anthropic researchers described how models trained to deliberately conceal certain motives from evaluators could still inadvertently reveal secrets, thanks to their ability to adopt different contextual roles or “personas.” The researchers were initially astonished by how effectively some of their interpretability methods… Read More »
The researchers observed this “emergent misalignment” phenomenon most prominently in GPT-4o and Qwen2.5-Coder-32B-Instruct models, though it appeared across multiple model families. The paper, “Emergent Misalignment: Narrow fine-tuning can produce broadly misaligned LLMs,” shows that GPT-4o in particular shows troubling behaviors about 20 percent of the time when asked non-coding questions. What makes the experiment notable… Read More »