Hello, and welcome to the latest AI Update. A special note for this edition: We have just launched a new site focused on AI security, along with a dedicated newsletter – Control Plane. If you enjoy ID Tech’s weekly AI Update, please go ahead and sign up for the newsletter, where I’ll be providing even more in-depth coverage of the AI security space.
Now, on to this week’s Update:
The White House has published a new framework urging Congress to create a national AI policy that would trump an emerging “patchwork of conflicting state laws” that “undermine American innovation and our ability to lead in the global AI race.” The framework brings with it policy recommendations that seem broadly concerned with softening a social backlash against AI, including policy objectives aimed at protecting children, shielding communities from rising electricity costs tied to data centers, and taking a harder stance against deepfake-driven fraud.
Microsoft is reportedly considering legal action over a partnership between OpenAI, in which it has invested billions, and Amazon Web Services. The latter firms have a set of agreements that will see OpenAI’s forthcoming enterprise platform, Frontier, offered through AWS. Microsoft would prefer to see that happen through its own cloud services, and claims exclusive rights to cloud provisioning of OpenAI’s API services. Speaking to the FT, an anonymous Microsoft source said, “We know our contract. We will sue them if they breach it. If Amazon and OpenAI want to take a bet on the creativity of their contractual lawyers, I would back us, not them.”
Nvidia revealed its Vera Rubin AI chip platform at its annual GTC conference this week, where CEO Jensen Huang revealed that the firm is anticipating $1 trillion in sales through 2027 thanks to “off the charts” demand for AI compute. The Vera Rubin platform essentially combines three key components – Rubin GPUs, Vera CPUs, and Groq 3 LPX units – to massively accelerate inference, which refers to the behind-the-scenes “thinking” that AI does in response to queries and prompts. Nvidia also unveiled “NemoClaw”, its own custom version of the OpenClaw agentic framework that brings additional security guardrails.
Jeff Bezos is in talks to raise $100 billion in a fund dedicated to revitalizing manufacturing with AI. The proposed “manufacturing transformation vehicle” would acquire various manufacturing companies, focusing on areas including chipmaking, defense, and aerospace, and would then use AI to speed up automation in those firms. The effort is seen as a sort of complement to Bezos’ previously announced Project Prometheus, a startup focused on leveraging AI in engineering and manufacturing applications.
World has launched a new developer tool aimed at tying the biometrically-verified identities on its platform to AI agents. Co-founded by Sam Altman, World has built a blockchain-based identity system that encourages users to verify that they are unique humans through an iris-scanning device called the “Orb”, while endeavouring to ensure that their personal identity data remains private. The new Agent Kit can be leveraged to cryptographically verify that a given AI agent is acting on the behalf of a verified, authorized human in various online transactions.
A co-founder of the major US data center provider Supermicro has been accused of violating sanctions against China by secretly selling tech to clients in the country. An indictment has been filed against Yih-Shyan “Wally” Liaw as well as two other company affiliates, alleging that they used a Southeast Asia-based intermediary, false end-user paperwork, repackaging, and staged “dummy” servers to hide the true destination of restricted AI systems. Among other shenanigans, they allegedly used hair dryers to remove and reattach labels and serial-number stickers to server boxes and dummy machines in order to help the fake inventory pass inspection.
We’re skipping the chatbot’s take this week (no offence, GPT-5.4!) to offer one of the interesting graphs from the inaugural Control Plane newsletter, which lays out the staggering rate of advancement in AI that remains invisible to most non-experts:
–
By Alex Perala
