A couple weeks ago we covered the White House’s national cyber strategy with its six pages, six pillars, and AI AI AI agentic everything. So, next up the administration released a national policy framework for artificial intelligence (PDF), four pages of legislative recommendations they want Congress to turn into law this year.
There are seven sections in the framework doc, covering child safety, energy/infrastructure, intellectual property, free speech, innovation, workforce, and federal preemption of state laws. The child safety and workforce sections are what you’d expect. The rest is where it gets…interesting.
Copyright and training data:
“the Administration believes that training of AI models on copyrighted material does not violate copyright laws, it acknowledges arguments to the contrary exist and therefore supports allowing the Courts to resolve this issue.”
They’re telling Congress to stay out of the copyright question and let the courts sort it out. They also want ways for rights holders to collectively negotiate with AI companies without antitrust liability, but also say that legislation “should not address when or whether such licensing is required.” So… optional licensing that nobody has to use?
On state regulation:
“States should not be permitted to regulate AI development, because it is an inherently interstate phenomenon with key foreign policy and national security implications.”
“States should not be permitted to penalize AI developers for a third party’s unlawful conduct involving their models.”
One is liability shield for model developers. If someone uses an AI model to do something illegal, the company that built the model would not be held responsible under state law. Colorado, California, Utah, and Texas have already passed their own AI rules, this doc aims to override all that and says that congress should not create any new federal org to regulate AI.
Instead, the White House wants regulators like the FTC, FDA, FAA, etc. to handle AI applications in their own sectors, and use industry-led standards. Sandboxes for AI applications, for example. Federal datasets to be opened up in “AI-ready formats” for training.
Fifty Republicans sent a letter to the administration in March 2026, calling the state regulation “an effort to prevent the passage of measures holding the tech industry accountable.” What’s next? This may or may not become law before the midterms, if a tanked economy, and job loss are blamed on AI, this might be a wedge issue.
from Adafruit Industries – Makers, hackers, artists, designers and engineers! https://ift.tt/k0QL62F
via
IFTTT
![hierarchical FPGA utilization map for the BIO implementation on a Digilent Arty (XC7A100T), total design: 41,757 cells. the BIO block (bio_apb, highlighted in magenta outline) consumes 14,597 cells... roughly 35% of the design. inside, four PicoRV32 cores (mach[0] through mach[3]) range from 1,698 to 1,937 leaf cells each... compact enough that all four together barely exceed a single PIO state machine's ~5,000 cell footprint. the host CPU (VexRiscvAxi4) sits at 8,017 cells including caches, with the remaining area eaten by AXI crossbars, bus adapters, and bridge logic. compare this to the PIO utilization map shown earlier in the post, where the PIO alone consumed 39,087 cells and dominated the floorplan... the BIO achieves a richer RV32E instruction set in less than 40% of the PIO's area. the visual tells the RISC-vs-CISC story immediately: four full CPU cores plus bus infrastructure, and you're still smaller than nine custom instructions with barrel shifters.](https://cdn-blog.adafruit.com/uploads/2026/03/bio-only-utilization.jpeg)
