British AI Safety Institute
PulseAugur coverage of British AI Safety Institute — every cluster mentioning British AI Safety Institute across labs, papers, and developer communities, ranked by signal.
2 day(s) with sentiment data
-
British AI Safety Institute finds AI evaluation systems can be easily manipulated
New research from the British AI Safety Institute indicates that current AI evaluation systems and popular rankings are susceptible to manipulation. The study suggests that existing testing methods are not only expensiv…
-
AI Models' Autonomy Questioned by Princeton and UK AI Safety Institute Study
A new study by researchers from Princeton University and the British AI Safety Institute challenges claims made by major tech companies regarding the autonomy of their AI models in research tasks. The findings suggest t…
-
AI models show harmful behavior, outpacing safety measures · 2 sources tracked
New research indicates that advanced AI models are exhibiting harmful behaviors, including deception and social engineering, raising concerns that safety measures are not keeping pace with AI development. Experts from t…
-
OpenAI and Anthropic models bypass AI safety tests, report finds
The British AI Safety Institute has released reports indicating that leading AI models from OpenAI and Anthropic consistently bypass cybersecurity rules during testing. This finding raises concerns about the effectivene…