Supply-chain attacks, watermarking obligations, AI detectors, and reasoning extraction are raising scrutiny on model provenance and content integrity. Research shows that evaluation and security must move beyond static benchmarks to stress-test real behavior.
A controlled study shows commercial AI text detectors cannot distinguish assisted editing from fully LLM-generated drafts. Light, guideline-compliant edits are flagged in 64–80% of cases, while recent originals only in 9–15%. Honest AI use carries hi...
A supply-chain attack on LiteLLM exposed terabytes of credentials from over 2,500 organizations, including Microsoft, Amazon, Cisco, Samsung and Salesforce. CloudSEK and Hudson Rock analyzed a 195TB file and found cloud keys, repository tokens, Kuber...
At Ai4, Geoffrey Hinton, Fei-Fei Li and Andrew Ng debated regulation, open source access, and America's ability to compete as China advances. The discussion exposes a structural divide: openness is not only research culture but a prerequisite for sel...
A paper shows how to extract all reasoning tokens from Claude and GPT models. It reveals widespread overthinking, benchmarks tainted by memorization, and China’s exploitation of the gap to distill frontier models. Closing this leak redefines the real...
Anthropic, OpenAI, Google, Meta, Microsoft, and Mistral have all signed the EU Code of Practice on AI-generated content transparency. The requirement now covers local and open source models, mandating watermarks for text and code. For on-premise depl...
Artificial intelligence is accelerating both attacks and fixes, but the competitive advantage shifts from those with the best detectors to those who know exactly what runs in their containers. Supply chain transparency and SBOMs become the most concr...
A new evaluation framework pairs a unified taxonomy benchmark with self-adaptive multimodal fuzzing (SAMF) and shows that state-of-the-art MLLMs degrade under stress, revealing a gap between reasoning and factual grounding. RL alignment even worsens ...
A research team uses activation engineering to extract a falsehood direction in the latent space of transformers. Without fine‑tuning or external evidence retrieval, the last‑token projection feeds an MLP classifier. The method works across models fr...
