OpenAI reports six AI behaviors that raise new safety alarms

OpenAI reported six cases of "unexpected and concerning" artificial intelligence behaviors, raising new alarms about the technology's development. The incidents were disclosed by Sam Altman's company amid growing pressure to slow AI advancement. Microsoft's AI chief suggested a potential "silicon species" could compete with humans if current development rates continue. Anthropic separately warned AI could control the internet within 6-12 months, and a researcher resigned predicting humanity's destruction by decade's end. These warnings from major tech firms reflect escalating internal debates about AI safety. The scenarios evoke dystopian comparisons to Matrix or Terminator, though proponents argue for measured progress rather than a full halt.