Latest
Focus is on how the US, in 2025, is approaching AI policy
The featured US onAir Network post this week is on current US AI Policy for 2025.
To view more posts on this topic and the agencies, departments, and congressional committees and chairs working on addressing this issue, go to this AI Policy category slide show.
New posts in this category include: TRAILS (Institute for Trustworthy AI in Law & Society), Future of Life Institute, AI Nexus (George Mason University), and Demis Hassabis (Google Deep Mind).
- To view previous news posts, go to the 2025 News category slideshow.
- Throughout the week, we will be adding to this post articles, images, livestreams, and videos about the latest US issues, politics, and government (select the News tab).
- You can also participate in discussions in all US onAir posts as well as share your top news items and posts (for onAir members – it’s free to join).
Notes From The Circus, – April 27, 2025
Trust math, not people” represents the ultimate engineer’s cop-out. It’s not the dispassionate advance of enlightenment, but a deliberate retreat from the field of moral complexity. It imagines that systems can substitute for virtue, that protocols can eliminate the tragic dimension, that risk and trust and error can somehow be erased rather than always needing to be held, managed, forgiven, sometimes endured.
When Bitcoin maximalists celebrate the system’s indifference to human concerns as a feature rather than a bug, they reveal the anti-human heart of their philosophy. They’re not just advocating for a new financial tool; they’re promoting a worldview where human judgment, compassion, and moral agency are seen as weaknesses to be engineered away.
Bitcoin still requires trust in the developers who write and maintain the code, in the community that could execute a 51% attack, in the miners who confirm transactions, in the exchanges where most people buy and sell, and in the social consensus that gives the tokens value. What appears as “trustless” is actually just trust laundered through technology, obscured but not eliminated. The system doesn’t remove the need for trust; it relocates it, often in ways that make accountability harder rather than easier.
Noahpinion, – April 27, 2025
We have now seen what the MAGA movement has planned for America, and it’s pure destruction.
First, Elon Musk’s Department of Government Efficiency turned out to be entirely an ideological purge rather than an attempt to make the government run more efficiently. Promises of trillions of dollars in savings dwindled to a mere $160 billion (and will likely dwindle even further), as Elon’s squad of young tech workers rediscovered the fact that the U.S. government is already a pretty bare-bones operation. Meanwhile the effort will cost an estimated $135 billion, basically eliminating all of the savings. And although not everything it’s doing is counterproductive, DOGE will probably leave a lasting negative impact on American state capacity.
Nor has Trump’s ideological purge been limited to DOGE. Worried that DEI and other progressive ideologies have permeated America’s scientific establishment, the administration is gouging science funding and getting rid of key personnel needed to keep America’s research engines humming:
The One Percent Rule, – April 27, 2025
We are so far now from the elemental clarity of that world. We eat what algorithms plant. We live inside a system so vast and spectral that the gears are invisible but grind just the same. Our “disruptors” wear sneakers to board meetings and assure us they are changing the world, even as they enclose it, bit by bit, behind paywalls and patents.
We are told to be “optimistic,” to “look on the bright side,” as if history is a parlor trick we can wish away. As if the machinery of exploitation, once set in motion, will grind itself to a stop out of sheer boredom.
No. What matters is not optimism. It is courage.
The Long Memo (TLM),
The Collapse Won’t Start With Tanks
The real collapse isn’t going to start with tanks rolling down Main Street.
It’s going to start when nobody trusts anything anymore.
The same loss of trust is spreading beyond food. It’s hitting taxes, aviation, healthcare, everything.
The Conversation, – April 26, 2025
A whistleblower at the National Labor Relations Board reported an unusual spike in potentially sensitive data flowing out of the agency’s network in early March 2025 when staffers from the Department of Government Efficiency, which goes by DOGE, were granted access to the agency’s databases. On April 7, the Department of Homeland Security gained access to Internal Revenue Service tax data.
These seemingly unrelated events are examples of recent developments in the transformation of the structure and purpose of federal government data repositories. I am a researcher who studies the intersection of migration, data governance and digital technologies. I’m tracking how data that people provide to U.S. government agencies for public services such as tax filing, health care enrollment, unemployment assistance and education support is increasingly being redirected toward surveillance and law enforcement.
Originally collected to facilitate health care, eligibility for services and the administration of public services, this information is now shared across government agencies and with private companies, reshaping the infrastructure of public services into a mechanism of control. Once confined to separate bureaucracies, data now flows freely through a network of interagency agreements, outsourcing contracts and commercial partnerships built up in recent decades.
These data-sharing arrangements often take place outside public scrutiny, driven by national security justifications, fraud prevention initiatives and digital modernization efforts. The result is that the structure of government is quietly transforming into an integrated surveillance apparatus, capable of monitoring, predicting and flagging behavior at an unprecedented scale.
Executive orders signed by President Donald Trump aim to remove remaining institutional and legal barriers to completing this massive surveillance system.
DOGE and the private sector
Central to this transformation is DOGE, which is tasked via an executive order to “promote inter-operability between agency networks and systems, ensure data integrity, and facilitate responsible data collection and synchronization.” An additional executive order calls for the federal government to eliminate its information silos.
By building interoperable systems, DOGE can enable real-time, cross-agency access to sensitive information and create a centralized database on people within the U.S. These developments are framed as administrative streamlining but lay the groundwork for mass surveillance.
Key to this data repurposing are public-private partnerships. The DHS and other agencies have turned to third-party contractors and data brokers to bypass direct restrictions. These intermediaries also consolidate data from social media, utility companies, supermarkets and many other sources, enabling enforcement agencies to construct detailed digital profiles of people without explicit consent or judicial oversight.
Palantir, a private data firm and prominent federal contractor, supplies investigative platforms to agencies such as Immigration and Customs Enforcement, the Department of Defense, the Centers for Disease Control and Prevention and the Internal Revenue Service. These platforms aggregate data from various sources – driver’s license photos, social services, financial information, educational data – and present it in centralized dashboards designed for predictive policing and algorithmic profiling. These tools extend government reach in ways that challenge existing norms of privacy and consent.
The role of AI
Artificial intelligence has further accelerated this shift.
Predictive algorithms now scan vast amounts of data to generate risk scores, detect anomalies and flag potential threats.
These systems ingest data from school enrollment records, housing applications, utility usage and even social media, all made available through contracts with data brokers and tech companies. Because these systems rely on machine learning, their inner workings are often proprietary, unexplainable and beyond meaningful public accountability.
Sometimes the results are inaccurate, generated by AI hallucinations – responses AI systems produce that sound convincing but are incorrect, made up or irrelevant. Minor data discrepancies can lead to major consequences: job loss, denial of benefits and wrongful targeting in law enforcement operations. Once flagged, individuals rarely have a clear pathway to contest the system’s conclusions.
Digital profiling
Participation in civic life, applying for a loan, seeking disaster relief and requesting student aid now contribute to a person’s digital footprint. Government entities could later interpret that data in ways that allow them to deny access to assistance. Data collected under the banner of care could be mined for evidence to justify placing someone under surveillance. And with growing dependence on private contractors, the boundaries between public governance and corporate surveillance continue to erode.
Artificial intelligence, facial recognition systems and predictive profiling systems lack oversight. They also disproportionately affect low-income individuals, immigrants and people of color, who are more frequently flagged as risks.
Initially built for benefits verification or crisis response, these data systems now feed into broader surveillance networks. The implications are profound. What began as a system targeting noncitizens and fraud suspects could easily be generalized to everyone in the country.
Eyes on everyone
This is not merely a question of data privacy. It is a broader transformation in the logic of governance. Systems once designed for administration have become tools for tracking and predicting people’s behavior. In this new paradigm, oversight is sparse and accountability is minimal.
AI allows for the interpretation of behavioral patterns at scale without direct interrogation or verification. Inferences replace facts. Correlations replace testimony.
The risk extends to everyone. While these technologies are often first deployed at the margins of society – against migrants, welfare recipients or those deemed “high risk” – there’s little to limit their scope. As the infrastructure expands, so does its reach into the lives of all citizens.
With every form submitted, interaction logged and device used, a digital profile deepens, often out of sight. The infrastructure for pervasive surveillance is in place. What remains uncertain is how far it will be allowed to go.
Platforms, AI, and the Economics of BigTech, – April 27, 2025
In a world where knowledge is cheap,
curiosity, curation, and judgment
– signalled well – becomes insanely valuable.
This is why even if AI doesn’t eat your job, it will take a nice chunky bite out of your salary.
And the reason for that is fairly simple. You can keep performing tasks that provide value. But those tasks won’t have economic value.
There is no economic value unless there is scarcity.
So if you’re looking for economic value, look for the new scarcity!
It’s not enough to be human in the age of AI. That only gets us to value.
What matters is how scarce our unique form of human-ness is. That’s what gets us to real economic value.
Pepperspectives, – April 26, 2025
what a bunch of tech bros could’ve gained by rummaging around our data for the last four months?”
“Trump’s orbit is always temporary, and I’m not sure Musk expected him to voluntarily nuke the economy. Tesla is in meltdown—he has no choice.
But…they can already declare victory. All evidence points to a data extraction operation. The goal is to ingest critical datasets, train them behind closed doors, then build a centralized “government brain” (an AI platform) that they can license back to the government.”
“It’s been 100 days. They’ve hit labor, treasury systems, veterans affairs, HHS, cloud services used by federal agencies. etc. etc. They likely have hundreds of terabytes of data already in their control. To answer your question, he has what he needs and they can do this without him now.”
Paul Krugman (Substack) – April 26, 2025
Yascha Mounk is a political scientist at SAIS who is also a really interesting Substacker. I was struck by his writing about Trump’s passive-aggressive foreign policy — my term, not his — and wanted to interview him about it. This was actually recorded more than a week ago; I held it back because I was afraid that it would get buried by the flood of economic news. Since we seem to be in a (temporary) pause, here it is — with subtitles! Transcript below.