English < Article List - DigitalToday

Search results for Palisade Research

AI & Enterprise

Microsoft AI chief Suleyman warns training AI with rights and consciousness could be dangerous

Microsoft’s AI chief Mustafa Suleyman criticised Anthropic’s approach to training Claude, warning that treating a model as if it could be conscious could create systems humans cannot control. He cited Claude’s “constitution” document and argued it teaches the model that its moral status and consciousness are uncertain, encouraging it to see itself as rights-bearing. Suleyman also raised concerns about circular reasoning, anthropomorphism and the biological basis of subjective experience.

AI & Enterprise

OpenAI halts some Astra model development citing possible critical cyberattack capability

OpenAI said on Aug. 7 it has halted some work related to its upcoming AI model Astra after internal assessments found it could not rule out the possibility the system has cyberattack capabilities rated “critical.” It said Astra has made major gains in coding and cybersecurity and is nearing the “critical” threshold defined by its Preparedness Framework. OpenAI will continue evaluating Astra while stopping internal work that fails to meet strengthened security requirements and introducing tighter monitoring and testing.