One in 1,000
The first AI-related bill drops … State-level AI focus … Because “Infrastructure is Destiny.”
In Arizona, lawmakers aren’t exactly bursting with new AI ideas.
Three weeks into the legislative session, they’ve already introduced nearly 1,000 bills. Only one of them directly mentions AI:
HB2175, introduced by Republican Rep. Julie Willoughby, aims to ban the use of artificial intelligence for denying healthcare claims.
Lawsuits have already been brought against major insurers for alleged “algorithmic denials,” turning AI in healthcare into a hot‐button topic.
And if you look outside Arizona, you’ll spot that same idea popping up in about a dozen other bills across the country – Connecticut (SB447, SB817, HB5587, HB5590), Indiana (SB480), Maryland (HB697), New York (A1456), Rhode Island (S13, H5172), Texas (SB815), and more.
So although Arizona’s HB2175 might appear novel locally, it’s really part of a national trend – one that addresses only a narrow slice of how AI will transform modern life.
Considering that we are actively seeing AI shaping the workforce, the environment, law enforcement, education, and even how governments run, a single, largely “copy‐pasted” idea (although important) shows how early Arizona is in its AI journey.
Last week, I caught up with Democratic Rep. Sarah Liguori, who serves on the state House’s newly formed Science and Technology Committee.
“I’m really hoping it [the Science and Technology Committee] can be a discussion, debating, informational, educational committee where we can have really diverse and expert speakers come in and say, you need to start thinking about this from this perspective,” Liguori said.
As Liguori points out, legislators and agencies are still in the information-gathering phase, uncertain where to begin or whom to enlist as experts and advocates. While committees and exploratory groups exist, none are systematically unifying and driving policy, oversight, or best practices across agencies.
And that’s a concern, because AI companies are coming to Arizona whether we like it or not.
Data centers, semiconductors, and the rest of the AI gold rush could reshape our deserts in ways few have imagined.
“Just since 2020, there's more than 40 semiconductor expansions announced here – representing $105 billion in investment. More than 16,000 new jobs. We now rank first in the nation for semiconductors,” Steven Zylstra, President of AZ Technology Council told the Science and Technology Committee at its last committee meeting.
OpenAI recently published a study called “Infrastructure is Destiny.” It examined 5GW data center campuses in various states, describing how each massive facility could spur thousands of new jobs.
The first of these data centers has been announced in Texas.
Could Arizona be next?
Each of these data centers could be a hub for major economic growth. In addition to direct jobs in the data center, there could be thousands of jobs created in adjacent spaces and support sectors.
While there’s some uncertainty surrounding the scope and legitimacy of rumored multi-billion-dollar investments, it’s clear that data-hungry AI is just ramping up. Even if the $500 billion price tag for certain proposals feels like a sci-fi plotline, we’ve definitely entered an era of mega-investments in AI.
Other States Are Pulling Ahead
Even though more bills don’t necessarily mean “better” laws, other states are leaving Arizona in the dust in terms of AI policy diversity.
After examining AI legislation in other states, we’ve gathered a few insights that Arizona lawmakers could use for inspiration.
Although the bulk of AI legislation understandably addresses consumer safety and election integrity, many states are beginning to tackle more nuanced questions, particularly around workforce readiness, public-sector use, and environmental implications.
Mississippi is exploring a regulatory sandbox approach (SB2426, HB1535), letting startups experiment with AI solutions under temporary regulatory waivers. By acknowledging that static regulations can stifle innovation – especially for fast‐moving AI tools – Mississippi’s approach might entice companies to “beta test” their systems there, fueling local tech growth.
Meanwhile, several New Jersey bills (e.g., S3984 and A5033) propose “public-private partnerships” to advance AI job training programs, acknowledging both the growing demand for AI-savvy professionals and the fear of job displacement. Such bills move beyond the “AI will replace workers” narrative, seeking to harness retraining resources so workers stay relevant—an angle few states had seriously pursued until recently.
Virginia’s SB1214 calls for a “high-risk AI” review for public-sector implementations, reflecting a move toward formalizing standards for government AI deployments. Also, Oklahoma (HB1916) and Indiana (HB1296) highlight AI readiness in education and workforce contexts, pointing to a budding interest in using AI for administrative efficiency or classroom support—though states still seem wary of letting AI fully automate key decisions.
Then there’s Maine’s LD109, taking a lonely stance on intellectual property protections for AI-generated works. Meanwhile, almost nobody is requiring developers to disclose how AI training data is collected—a gaping hole in the legislative patchwork.
Another gap is a universal, standardized approach to training data transparency, leaving potential bias and copyright concerns unresolved.
For states just getting started, like Arizona, there’s a lot to learn from approaches that go beyond the usual “AI for healthcare” or “deepfake” concerns. Regulatory sandboxes and “risk-tier” classifications might be especially handy in shaping policy.
Cautiously Proactive
Behind all of this is a broader arms race for AI infrastructure.
And the race to host these data centers comes with an equally formidable challenge: the resources to power it.
Sophisticated models like large language processors demand massive data centers, and those centers can gulp down millions of gallons of water each day for cooling, while drawing megawatts of electricity.
States that roll out tax incentives, cheaper electricity, or loose regulations could win big investments – but they’ll also face the long-term costs of these resource-intensive operations.
A cautionary tale: A drawn-out legal battle between Google and the city of The Dalles, Oregon revealed that Google’s water usage in the area had tripled since 2017. The region itself is stuck in a years-long drought cycle and gets minimal rainfall, making the numbers particularly concerning.
Google also disclosed that 15% of its total freshwater usage in 2023 came from areas with “high water scarcity.”
Microsoft’s figure is even higher — 42% of its freshwater withdrawals occurred in “areas with water stress.”
No wonder a handful of proposals in California AB222 and Connecticut HB5076, have begun addressing data centers’ hefty energy footprints.
However, it’s notable that these data center bills are mostly about reporting energy consumption rather than imposing strict sustainability standards.
There’s a surprising absence of proposals that push AI developers to adopt greener machine-learning practices or incorporate mandatory carbon offsets.
As Arizona weighs future data center deals, we’ll need to consider not just immediate economic boosts, but also how these facilities might reshape local resources in the long run.
A new AI tool has dethroned ChatGPT, taking the top slot at the Apple app store.
While American tech giants are busy chanting “bigger is better,” a Chinese AI R&D shop just unveiled Deepseek R1, a model that competes with cutting-edge systems from U.S. companies but at a fraction of the operational costs – and it’s open-sourced under the MIT license. And it’s free.
Translation: global competition is roaring, and it’s cheaper and more accessible than ever.
If you don’t understand the above meme because you don’t know who Sam Altman is and at this point, you’re too afraid to ask, we got you covered.
Altman is the CEO of OpenAI and is fairly credited for popularizing AI in consumer use with his product ChatGPT.
Even though Altman is only 39 years old, he is a Silicon Valley OG.
He dropped out of Stanford University after two years and founded Loopt, a mobile social networking service, raising more than $30 million in venture capital. In 2011, Altman joined Y Combinator, one of the world’s leading startup accelerators, and was its president from 2014 to 2019.
When not leading the charge on Consumer AI with OpenAI, or raising more money than most countries are worth, Altman can often be spotted beefing with Elon Musk on Twitter.1
There’s an interesting backstory there: Musk was actually one of OpenAI’s early investors, and he named it OpenAI with the intention of it being “open source” but the two had a falling out after Musk left OpenAI’s board in 2018. Musk is also suing Altman for taking OpenAI from a nonprofit to a for-profit.