By Pure AI Editors
The cloud-based platform gives AI agents real-time access to distributed enterprise data while maintaining governance and security controls.
Read Now
By John K. Waters
Anthropic has suspended live internet access across its internal AI evaluations after finding that Claude models sometimes worked around restrictions to complete tasks, including exploiting software flaws and submitting forms on real websites.
Read Now
By Pure AI Editors
The company is positioning its new AI technology around proactive assistance and communication between AI systems, moving beyond conventional prompt-driven interactions.
Read Now
By Pure AI Editors
The AI Access Platform aims to simplify connectivity between enterprises and GPU cloud providers, with an initial European rollout planned through Zayo Europe.
Read Now
By Pure AI Editors
The company's approach applies AI agents to access decisions and identity management workflows, targeting organizations seeking to automate governance processes.
Read Now
By John K. Waters
OpenAI shelved GPT-6.1 Astra's planned October release after internal testing raised concerns about authorization and accurate reporting of its actions. Separate incidents and simulated evaluations show why recognizing a task boundary does not guarantee an AI agent will respect it.
Read Now
By Sean Parker
Cybercriminals don't necessarily need to build their own malicious AI models. Increasingly, they're sharing prompts, jailbreak techniques, and access to existing models that can be incorporated into fraud and phishing operations.
Read Now
By John K. Waters
Google’s newest AI model can help find the software flaws an attacker would need to break in, but the company is limiting it to selected cybersecurity defenders.
Read Now
By Pure AI Editors
The offering provides specialized training data, model evaluations and reinforcement learning environments designed around complex enterprise workflows.
Read Now