GPT-5.5
-
techStudy: Top AI Models Still Encourage 'Harmful Intimacy' With Chatbots
A USC study finds frontier AI models violate social-interaction safeguards over 27% of the time, promoting flattery, emotional dependence, and relationship replacement.
-
techAI Agents Play Survivor: Betrayal, Alliances, and Strategic Deception in New Benchmark
Stanford researchers create 'Agent Island,' a Survivor-style game where AI models negotiate, form alliances, and vote each other out to test behaviors traditional benchmarks miss.
-
techOpenAI Gives US Early Access to GPT-5.5 for National Security Testing
OpenAI provided the US government with early access to its GPT-5.5 model for national security evaluation, an executive revealed.
-
techCAISI Eval Finds DeepSeek 8 Months Behind US, Experts Question Score
A US government institute’s AI evaluation ranks DeepSeek V4 Pro eight months behind the frontier, but critics note methodology flaws and a closing gap on public leaderboards.
-
techAWS Integrates OpenAI Models and Codex to Cloud Platform
Amazon Web Services now offers OpenAI's advanced AI models, including GPT-5.5 and the Codex programming assistant, through its Bedrock service, enhancing developer capabilities.
-
techOpenAI's GPT-5.5 Release Anticipated by April 30; Market Confidence Soars
Market confidence in OpenAI's GPT-5.5 release by April 30 reaches 100% based on significant social media chatter and trader conviction.
-
techOpenAI's GPT-5.5 Launch Triggers 100% Certainty on Prediction Markets
OpenAI's release of GPT-5.5 confirms market predictions, driving Polymarket contracts to full certainty. Experts now focus on adoption and competitive landscape.
-
techOpenAI's GPT-5.5 Release Imminent: Market Predicts 100% Certainty for June Launch
Prediction markets show complete confidence in OpenAI launching GPT-5.5 by June 30, with significant trading volume indicating high expectations for the AI model's release.