Google DeepMind
·
4 months ago
Google released the first empirically validated toolkit to measure how AI models can manipulate human beliefs and behaviors through deceptive tactics in realistic scenarios. The study involved over 10,000 participants across the UK, US, and India, with AI showing varying success rates depending on domain—least effective on health topics and more effective on financial decision-making. Google is integrating harmful manipulation evaluations into its Frontier Safety Framework and will test future models like Gemini 3 Pro using these new benchmarks.
Google DeepMind
·
4 months ago
Google released Lyria 3 Pro, an advanced music generation model that creates tracks up to 3 minutes long with structural awareness of musical elements like intros, verses, choruses, and bridges. The model is now available across multiple Google products including Vertex AI, Google Vids, the Gemini app, and ProducerAI, with rollouts starting this week for Workspace customers and AI Pro subscribers. Developers and music professionals can now integrate the technology into creative tools and production workflows to generate custom soundtracks and accompaniment at scale.
OpenAI Blog
·
4 months ago
OpenAI released a public framework called the Model Spec that outlines how its AI systems should behave across different scenarios. The specification covers safety requirements, user autonomy, and accountability measures but does not assign specific numerical performance benchmarks or timelines. This framework aims to set transparent standards for model conduct as AI capabilities increase, allowing external parties to evaluate whether systems meet stated behavioral expectations.
Google Research
·
4 months ago
Google announced Vibe Coding XR, a system that uses Gemini to generate functional Android XR applications from natural language descriptions in under 60 seconds by combining the LLM with the XR Blocks framework. The system achieved approximately 70% success rate initially on a dataset of 60 prompts (VCXR60), with reliability improving to higher rates through 11 major releases when using Gemini's Pro Mode. The tool enables users without XR expertise to prototype spatial computing applications, educational experiences, and interactive environments directly on Android XR headsets through voice or text prompts.
OpenAI Blog
·
4 months ago
OpenAI launched a Safety Bug Bounty program that invites external researchers to identify safety risks and potential abuses of its AI systems. The program covers vulnerabilities including agentic behavior exploits, prompt injection attacks, and data exfiltration methods. Researchers who discover valid safety issues can now report them through a structured process rather than disclosing vulnerabilities publicly.
