Midjourney’s ultrasound demo raises validation questions
Midjourney released a behind-the-scenes video of a dunk-tank ultrasound scanner and announced spa rollouts – but clinical proof and regulatory clarity are missing.
Mark Zuckerberg says Meta’s AI agents are behind schedule
In an internal meeting, Mark Zuckerberg acknowledged that work on Meta’s AI agents hasn’t advanced as quickly as he expected – a rare sign that the company’s flagship AI push is slowing.
Anthropic’s Fable 5 cleared to return after political review
After weeks of negotiation with the US administration, Anthropic will restore Claude (Fable 5) across major clouds. The decision reframes model rollouts as politically negotiated products.
Libby will filter out AI content, kind of
OverDrive’s Libby plans a filter for AI-made books that leans on publisher flags and metadata rather than reliable automated detection – shifting control back to publishers.
Palantir Brings Nemotron Open Models to US Agencies
Palantir announced an intelligence engine that pairs NVIDIA’s Nemotron open models with Palantir’s hardened, isolation-first deployments for U.S. agencies – packaging open LLMs into accredited, closed environments.
Zhipu AI (Z.ai) says GLM-5.2 matches Mythos on specific cybersecurity tasks
Zhipu AI released open-weight GLM-5.2 and early tests claim parity with Anthropic’s Mythos on bug-finding and exploit-surface discovery – a focused, not blanket, capability shift that raises deployment and policy questions.
Suno’s Spark incubator blurs artist support and data capture
Suno launched Spark, an incubator that pays grants and promotes unsigned artists while routing their output into the company’s AI and streaming pipeline. This is as much a content-acquisition play as a talent program.
Trump admin clears Anthropic’s Mythos 5 for 100+ U.S. companies and agencies
The U.S. government has authorized more than 100 companies and agencies to use Anthropic’s Mythos 5 – including non-U.S. employees – a move that speeds LLM adoption and raises immediate cross-border privacy questions.
Anthropic’s Mythos takedown signals regulatory shift
Two weeks after a government ultimatum, Anthropic pulled its Mythos models offline. The move is a clear signal: regulators will act fast, and access to frontier models may hinge on compliance, not just capability.
DeepSeek Reportedly Raises $7.4 Billion at >$50B Valuation – Pulse 2.0
DeepSeek has reportedly closed a $7.4B private round at a valuation north of $50B. The deal is a market signal: institutional capital is again concentrating on category leaders, with consequences for hiring, infrastructure and smaller startups.