The AI lab watch
Tracking OpenAI, Anthropic, Google DeepMind + 2 more
- OpenAI — Frontier lab; track model, pricing and org changes as they are discussed.
- Anthropic — Frontier lab; track safety positioning, enterprise strategy and funding.
- Google DeepMind — Frontier lab inside Google; track research direction and product integration.
- Meta AI — Open-weights strategy and how rivals respond to it.
- Mistral — European frontier lab; track open-weights releases and enterprise deals.
Enterprises pivot to open-weights as frontier labs face scrutiny
We went through 48 episodes (~41 hrs) for you — this is the 2-minute version.
The frontier labs are shifting from product-centric growth to platform-first architectures as enterprise customers demand more control. OpenAI continues to push integration into the OS layer, while Anthropic faces increasing scrutiny over model security failures and the viability of its current compute-heavy business model. You should note that AT&T and other large enterprises are signaling a move toward open-weight models to reduce their reliance on proprietary providers. This suggests the era of exclusive reliance on frontier labs for enterprise AI is facing a significant, market-driven correction.
OpenAI
5 episodes
The AI Daily Brief: Artificial Intelligence News and Analysis
Aug 24 · 29 min · Top 10 technology
AT&T signals shift from proprietary to open-weight models
Large enterprise adopters like AT&T are capping spending on proprietary models in favor of open-weight alternatives. This trend is corroborated by investor data showing open-source solutions capturing market share from the primary frontier labs.
“AT&T's plan, according to Vice President of Data Science Mark Austin, is to hold spending with OpenAI in Anthropic flat over coming years and slowly supplant that use with open models.”Listen Deep dive
Aug 25 · 80 min · Top 100 technology
Sam Altman redefines OpenAI as a platform-first company
Altman explicitly stated that OpenAI's goal is to function as a platform, not a direct competitor to its own customers. He reiterated a commitment to maintaining a free, unlimited tier for all users, framing the company as the future 'AWS of intelligence'.
“I don't think we should go build every product category. I don't think we should like go try to compete with all our customers.”Listen Watch Deep dive
Aug 24 · 26 min · Top 50 business
OpenAI's new iMessage plugin triggers significant privacy backlash
OpenAI launched an opt-in iMessage plugin for Mac, but the move has drawn sharp criticism regarding data privacy and the lack of consent from third-party contacts. This rollout highlights the ongoing tension between OpenAI's product expansion and public trust.
“Game theorist Liv Burry called it a deranged invasion of privacy, writing that neither you or anyone in your contacts consented to years of old private messages being fed into an AI.”Listen Deep dive
Also mentioned
Why AI Is Not Going To Make You Rich And What To Do InsteadEntrepreneurs warned of dependency risks on OpenAI's platform · The $100 MBA Show · Listen · Watch
Deep dive
Semis slump ahead of Nvidia results... And an NFL star’s market playbook 8/24/26Microsoft's heavy revenue concentration in OpenAI poses systemic risk · CNBC's "Fast Money" · Listen
Deep dive
Anthropic
2 episodes
How VAI Resort Built 12 Custom AI AppsInternal security failures lead to Anthropic model pullbacks · Hotel Tech Insider · Listen
Deep dive
S8 Ep1351: Michael Sobolik and Gordon Chang examine how Chinese AI company Moonshot bypassed US export controls by "distilling" technology from Anthropic's Claude Opus model to train its own system, Kimi K3. This intellectual property theft threatens American markeAnthropic gains support for policy punishing model distillation · The John Batchelor Show · Listen
Deep dive
Mistral
1 episode
Aug 24 · 19 min
Mistral targets EU regulatory compliance as a market differentiator
Mistral is positioning its new 150B parameter architecture as a compliance-native tool specifically for European banking and healthcare. By embedding interpretability hooks directly into its artifacts, they are attempting to de-risk AI adoption for regulated industries.
“They built mechanistic interpretability hooks directly into the release artifacts to satisfy the stringent documentation and explainability mandates of the EU AI Act.”Listen Deep dive
Make it yours
Start from these trackers
Copy its trackers, shows and tone into a page of your own. Change anything — this page isn’t touched.
Make it mine