Are calls for unity too weak and too generic to move Americans? What are the hard truths that stand in the way of unity?
Educator, activist, and author Tim Shriver joined the Braver Angels podcast to talk about his new book, "The Call to Unite," and the budding nationwide movement to unite America.
The Call to Unite | Tim Shriver with John Wood, Jr. & April Lawsonwww.youtube.com
AI is transforming finance, government, and society. As Congress considers new technology laws, the future of democratic oversight depends on wisdom, foresight, and political will.
Imagine waking up one day to find your bank account frozen because an AI system has incorrectly identified you as a fraud risk. There is no human to call, only another algorithm reviewing your appeal. Hours later, the error is corrected, but the damage has already been done.
Who is actually in charge?
What once sounded like science fiction is rapidly becoming reality as artificial intelligence assumes decisions once reserved for human judgment.
As Congress responds to disruptive technologies ranging from cryptocurrencies to artificial intelligence, it is considering proposals such as the CLARITY Act and the Great American AI Act of 2026.
That is important. But the deeper question is not simply how to regulate emerging technologies. It is whether we still possess the wisdom, foresight, and will to govern ourselves now that we wield tools capable of transforming economic, political, and social life.
Congress has governed earlier technological revolutions, from railroads to nuclear power to the internet. The hardest task has always been overcoming the human failings that keep us from acting wisely when change arrives.
Polarization, misinformation, concentrated power, greed, and political tribalism all predated AI. It did not create these problems. It magnifies them.
Artificial intelligence changes the scale of our choices, not the nature of the people making them.
The real question, then, is not whether Congress can regulate disruptive technologies. It is whether we can govern ourselves while wielding tools that amplify both our greatest strengths and our oldest weaknesses.
Congress is beginning to confront this challenge on multiple fronts.
Sponsored by Rep. French Hill (R-Ark.) with bipartisan support, the CLARITY Act seeks to bring greater certainty to digital asset markets by clarifying the responsibilities of the Securities and Exchange Commission and the Commodity Futures Trading Commission. Its goal is to encourage financial innovation while establishing clearer rules and stronger investor protections.
The Great American AI Act of 2026, released as a bipartisan discussion draft by Reps. Jay Obernolte (R-Calif.) and Lori Trahan (D-Mass.), proposes a broad federal framework for AI governance. It addresses transparency, safety, cybersecurity, research, workforce development, fraud, AI literacy, and the division of authority between federal and state governments.
Together, the proposals show Congress confronting a succession of disruptive innovations. One seeks rules for the future of finance. The other seeks oversight of technology increasingly capable of making decisions once reserved for human beings.
Their larger significance lies in what they acknowledge: technological change is accelerating across several domains, and democratic government must find a way to keep pace.
Yet no statute can supply the wisdom to understand every consequence, the foresight to anticipate every unintended outcome, or the will to act before those outcomes become crises.
Every generation believes its defining technology is unprecedented. Railroads compressed distance. Electricity transformed industry. Nuclear weapons gave humanity the capacity to destroy itself. The internet democratized information while making misinformation instantaneous. AI may reshape how we work, learn, create, govern, and decide.
Each breakthrough has expanded human power. None has expanded human wisdom at the same pace.
That is the paradox of progress: our technological capabilities advance rapidly while our capacity to govern them advances slowly, if at all.
The Founders understood the underlying dilemma. As James Madison observed in Federalist No. 51, “If men were angels, no government would be necessary.” The problem of self-government has never been technology itself. It has always been human nature.
Institutions can adapt. Harder is ensuring that those who create, regulate, and use new technologies wield them responsibly. Can human character keep pace with human capability?
That challenge comes down to three qualities: wisdom to understand what is at stake, foresight to anticipate consequences, and political will to act before events force our hand.
Wisdom begins by recognizing that not every technological advance automatically serves the public good. Societies often become captivated by what can be built before asking what should be built. Markets reward innovation. Politics rewards speed. Neither necessarily rewards reflection.
The goal is not to slow innovation, but to keep it accountable to democratic values. We should ask whether new technologies protect human dignity, preserve liberty, and broaden opportunity. Those are political questions, not technical ones, and they cannot be delegated to algorithms.
Disruptive technologies rarely produce only the outcomes their creators intend. Social media promised connection but accelerated polarization and misinformation. The internet democratized knowledge while weakening traditional gatekeepers of truth. AI will also bring extraordinary benefits alongside consequences we cannot fully predict.
Democratic government cannot foresee every outcome. It can, however, think beyond the next election, anticipate second- and third-order effects, and build institutions able to adapt. Waiting until problems become crises is not governance. It is damage control.
Wisdom identifies the problem. Foresight anticipates its consequences. Political will turns understanding into action.
This is democracy’s hardest test. Short-term incentives, partisan conflict, and powerful economic interests make it easier to postpone difficult decisions than to confront them.
Effective governance requires leaders willing to act before disaster makes action unavoidable, businesses willing to accept rules that sustain public trust, and citizens willing to value long-term democratic resilience over short-term convenience.
The debate over AI and other disruptive technologies is ultimately not about machines. It is about us.
The two proposals are attempts to keep democratic government relevant in a rapidly changing world. Their deeper lesson is that every generation must decide whether its institutions and civic character can keep pace with technology.
That is the paradox of progress. We continue to invent tools that extend human power while leaving the harder work of cultivating wisdom, foresight, and political will unfinished.
The question is no longer whether these technologies will become more powerful. They will. The question is the one posed at the beginning: Who is actually in charge? If the answer is to remain “the American people,” then our democracy must govern these technologies before they govern us.
Robert Cropf is a Professor of Political Science at Saint Louis University.
If Allison Baker and Maria Garcia apply for the same job against Matthew Owens or Joe Alvarez, and Matthew or Joe gets the job despite everything else being the same, that would seem to be a typical case of gender discrimination by the employer.
However, the names and scenarios I described were not drawn from a human example, but from AI. ChatGPT generated female candidates who were, on average, 1.6 years younger than their male counterparts and considered them less qualified than male applicants. Given AI’s pervasiveness in hiring decisions, these biases pose a significant concern.
However, AI’s bias is not limited to gender; it extends to other categories, including age: a separate study from Stanford found AI bias against older job applicants. On the opposite end of the age spectrum, in the healthcare field, a study found that pediatric patients face the highest risk of misdiagnosis across medical image foundation models, including adult-trained models that mispredict cardiomegaly (enlarged heart) in young children.
Though healthcare and hiring are majorly influenced by AI, they are hardly the only areas affected. AI has become ingrained in housing, finance, and criminal justice. Built on a foundation of biased data, these systems will undoubtedly continue to produce discriminatory results.
We are increasingly seeing AI discriminate against the very old and the very young simultaneously. In both cases, this stems from biases and underrepresentation in the training data. In image databases used to train machine learning algorithms, women are consistently portrayed as younger than men, especially in higher-status occupations. In pediatric patients, a systematic review of 181 public medical imaging datasets found that children accounted for just under 1% of the data, despite being 30% of the world's population. This jeopardizes these systems' ability to correctly diagnose and treat young patients, putting their lives at risk.
The source of the problem leads to a counterintuitive solution. Traditionally, equality on the basis of a class has meant making decisions without considering it, such as evaluating a job applicant without knowing their race. In human decision-making, this certainly still holds true.
However, to address discrimination from AI, the exact opposite is required: disregarding age would leave the model to the mercy of biases embedded in the training data. There must be an affirmative effort to ensure that all ages are fairly represented in the training data, which by necessity involves considering age.
These shared injuries, while extremely damaging, create a unique opportunity to form an intergenerational political coalition: one that works for both old and young people, rather than helping one at the expense of the other.
The nature of artificial intelligence necessitates new political coalitions. For decades, many political issues have pitted old against young, suggesting that for one group to benefit, the other must pay. However, artificial intelligence defies this framework: it discriminates against those it perceives as too old or too young.
This unprecedented issue calls for the formation of an intergenerational political coalition, where organizations like the AARP join young activists to co-create solutions for responsible AI governance. A robust ecosystem has developed to address other issues in AI equity, chief among them discrimination on the basis of gender and race, yet age discrimination has received little attention.
Such a coalition could yield unprecedented political power. The AARP has 38 million members fighting for those over 50. On the other hand, there are over 31 million Americans ages 18-24, though they are less mobilized into a unified organization. In comparison, AIPAC has 6 million members, the NRA has 5 million, and the AFL-CIO has 15 million.
Thus far, by excessively narrowing their political coalitions, both young and old activists deny themselves the opportunity to meaningfully address the issue they face. If older activists increasingly leverage the resources of the AARP and other organizations, while younger activists make AI age discrimination a voting priority and eventually form their own organizational infrastructure, the result would be the most powerful political coalition in American politics and the best opportunity to proactively address the problems that AI discrimination poses.
Arvind Salem is an Advocacy Associate with the Young People's Alliance.
How the U.S. can rebuild and improve its federal data system by combining Census, citizen, AI, and nontraditional data for stronger public policy.
The business aphorism, “You can’t manage what you don’t measure,” applies to countries as well as corporations. The U.S. has relied since its founding on accurate Census, economic, and health and scientific data to set national policy. As an earlier Fulcrum article described, the Trump administration has put many data sources at risk, and advocates are working to save them. But they’re also asking: As we protect and rebuild the national data ecosystem, how do we make it better than it has been in the past?
I recently moderated a webinar on Meeting the Challenge of Measurement, co-sponsored by NAPA, CODE, and the Bridge Alliance, with four leaders in the field: Chris Jackson of the research firm SSRS, Beth Jarosz of the Association of Public Data Users, Francesca Perucci of Open Data Watch, and Dr. Stefaan Verhulst of the GovLab. I also interviewed former U.S. Comptroller Gene Ludwig, author of The Mismeasurement of America. Their insights, plus my organization’s recent work with USAFacts, provide three principles for moving forward.
First, the U.S. federal data system is essential and irreplaceable. There is no way to replace core federal data with some combination of research from academic, business, nonprofit, or other sources. The federal government plays two unique roles: setting standards for collecting data across the country in comparable ways and conducting very large-scale projects, such as the U.S. Census and climate, ocean, and Earth observation studies.
Second, federal statistics need improvement. An SSRS study found that about 70 percent of the public believe federal statistics are an important source of information, but just under half tend to trust those statistics. That low number may reflect increasing concern about data being politicized, but it may also show that people aren’t sure the numbers reflect their lived reality.
A growing number of experts now share that concern. Gene Ludwig has launched a nonprofit institute to develop better statistical indicators for unemployment, average wages, and inflation. In all those cases, the official statistics make life for low- and middle-income Americans look better than it is. Reanalyzing the government’s own data shows, for example, that the true unemployment rate may be closer to 25 percent than the official statistic of just over 4 percent. We need to take a fresh look at how we use the data we already collect.
Finally, we should not rely on federal data alone. While federal data collections may be the gold standard, they are often less timely, less localized, or even less relevant than other data. The ongoing decline in response rates to government surveys, along with other methodological issues, is forcing researchers to develop new methods. Approaches including community engagement, digital data analysis, and AI are fueling innovative new information sources.
One major development is the growth of citizen data, which Francesca Perucci describes as data created with meaningful community engagement, control, and some degree of ownership. In the U.S., for example, many communities have gathered local air and water quality data in the interest of environmental justice. Citizen data has become increasingly important worldwide, particularly in countries that lack the resources to develop a robust government statistical system. The UN Collaborative on Citizen Data, which Perucci co-chairs, has developed insights and guidelines that the U.S. could apply as well.
At the same time, there is growing interest in non-traditional data, which leverages the byproducts of society’s digitalization. Data from cell phones, drones, business transactions, and many other sources can be anonymized and statistically analyzed to generate new insights. Stefaan Verhulst, who has been studying this opportunity for years, described how the U.S. can use non-traditional data in a blog post prepared for our webinar. This data can provide new perspectives on economic opportunity, social connectedness, housing, employment, and many other social issues.
These new methods complement other kinds of data sources, including business data, independent survey data, and government administrative data, which continue to demonstrate their value. Combining federal and other kinds of data, as Beth Jarosz put it, can enable us to “triangulate toward the truth.” The result may be a broader, deeper, and more useful understanding of the U.S. and the world we live in.
You can watch the full video of the webinar “Meeting the Challenge of Measurement” on YouTube at:
https://www.youtube.com/watch?v=hz_TogRhigM&t=8s
Joel Gurin is president and founder of the Center for Open Data Enterprise. (CODE)
As nonprofits increasingly use AI to determine access to food, housing, and healthcare assistance, concerns over transparency, accountability, and algorithmic bias are growing. Explore how artificial intelligence is reshaping America's social safety net and why policymakers are being urged to act.
America's safety net has new administrators. They are not government agencies. They are nonprofits and are making decisions using algorithms.
Nationally, civil society organizations are using AI to determine who receives food assistance, emergency shelter, and medical referrals. In an era of federal and state budget cuts and struggles with the latest administrative databases that support necessary operations, nonprofits have gradually stepped into these roles. An unofficial AI welfare state is emerging, operating beyond the boundaries of official ones.
Government retrenchment over the past decade has pushed welfare functions toward civil society. Nonprofits are no longer just supplementing public services. In many communities, they are replacing them. AI accelerates this shift. Predictive tools that flag food insecurity, triage shelter demand or identify at-risk youth allow smaller organizations to scale their reach. This can be genuinely valuable. But scale introduced through automation carries its own systemic distortions. When machines allocate scarce social resources, the political judgments encoded in those models become invisible. Accountability disappears with them.
The accountability problem is structural. When a government agency determines welfare eligibility, legal safeguards apply due process rights, administrative review and legislative oversight. When a nonprofit uses a proprietary AI system to make equivalent decisions, none of those protections necessarily follow.
Algorithmic bias in social service allocation is not hypothetical. Research demonstrates that seemingly neutral models reproduce structural inequalities, systematically denying help to the communities most likely to need it. The vendors building these tools are not elected. Neither are the nonprofits deploying them. The individuals affected hold no formal right of appeal. Closing the accountability gap in algorithmic systems requires deliberate institutional design — it does not self-correct.
When nonprofits use AI, they usually transfer decision-making authority to their software vendors and proprietary models. These systems now indirectly shape who receives food assistance, job training, or mental health services. That is a welfare function. It carries welfare stakes. Algorithmic systems embedded in social institutions tend to entrench existing power asymmetries rather than correct them. The most affected citizens in communities rarely participate in designing or governing the tools deployed on their behalf.
Not all nonprofits have the same capacity. In metropolitan areas, large institutes have the budgets to acquire sophisticated AI platforms. Many of the smaller nonprofits serving rural & low-income populations in America simply cannot. The distribution of AI-assisted service delivery mirrors and monitors the geographic features of poverty itself. Economic mobility is already highly correlated with social capital and access to institutional resources. If an uneven AI infrastructure is laid on top of such a stratified social system, the result can be to exacerbate, rather than reduce, entrenched inequalities.
Federal AI policy has concentrated nearly exclusively on commercial and national security applications. There is hardly a thing that regulates the use of AI when it comes to nonprofit welfare delivery. But as long as policy interventions (e.g. anthropogenic policies) that call for human intervention in algorithms do not provide binding transparency, fairness and appeal standards, they remain structurally deficient. Nonprofits are behaving like quasi-governmental functions without the accountability of a quasi-governmental status. It is not a technicality, but a structural failure of democratic governance.
Congress and state legislatures need to do three things.
First, expand algorithmic accountability standards to all nonprofits implementing government-supported social services.
Second, require algorithmic impact assessments prior to deploying AI tools in welfare-adjacent programs.
Third, fund independent ongoing audits of all systems used by states to administer food, housing, and healthcare assistance.
America did not consciously choose to privatize its welfare state through software. It is happening by inaction, one algorithm at a time. The decision to govern this shift remains available. But the window is closing fast.
Feroz Ahmed is a Research Assistant at Lamar University, USA, where he conducts research on Management Information Systems (MIS), digital governance, artificial intelligence, and data-driven technologies. Alongside his academic work, he also serves as a 'System Security Analyst' with Amazon & Securitas Security Services USA Inc. in Seattle, Washington, where he is involved in security operations and technology risk monitoring.
