Asana's AI agents share memory across your company — but not your secrets
THE SO WHAT
Shared memory for agents—without leaking sensitive data—is the difference between one-off copilots and a real organizational nervous system. If you’re piloting agents, the question now is less “can it answer?” and more “what does it remember, how is that scoped, and who controls the retention rules?”.
READ THE SOURCE
MORE FROM THE WIRE
Applied AIDisrupting a Criminal Scam Operation
AI-native fraud is now organized enough that labs are running counter-intel ops against specific scam shops. If your product touches payments, dating, or gambling, assume you’re already in an arms race with LLM-augmented fraud rings and budget for continuous abuse tooling, not one-off rules.
Applied AIAfter killer quarter, Palantir CEO Alex Karp calls AI industry ‘Marxist’
Enterprise buyers are being told—again—that frontier labs are too misaligned and opaque to trust with sensitive workloads, just as Palantir reports $1 billion in profit. Expect more boards to ask whether their AI stack is built on vendors whose incentives, governance, and security posture they actually understand.
Hugging Face CEO says China is winning the AI race while the US is building 'in silos'
The argument that open-weight models are now a defense asset reframes “open vs closed” as a national capability question, not just a licensing debate. If you’re in US or allied markets, expect more pressure—from regulators and customers—to justify closed stacks that don’t interoperate or allow independent security review.
Applied AIA DeepMind exec finally said what the trillion-dollar AI spend is for: machines that improve themselves.
If the trillion-dollar capex story is about recursive self-improvement, then today’s models are just bootloaders for systems that tune their own weights, data, and tools. For operators, that means planning for AI that changes faster than your current governance and validation cycles can comfortably handle.