This Storyboard - which we call our "stain" chart - shows you at a glance how strong or weak a given narrative is right now relative to its history.

For each narrative or "semantic signature" listed on the left of the chart, we have a series of blue dots on the right, each of which represents a specific weekly density or volume of that narrative. reading from within the date range that we are covering. The red arrow is the most recent reading, so it's just like the "YOU ARE HERE" spot on a map. The x-axis scale shows the range of index values.  If a dot is at 100, that means that story is 100% more present in media than usual. If it’s at 0, it means it’s at its normal level.

The light blue shaded box covers the middle 50% of readings across the date range, so you can see quickly if the current reading is typical (inside the blue box), depressed (left of the blue box), or elevated (to the right of the blue box).

If you hover over a specific blue dot, you will see the specific date and measurement that the dot represents.

The Pulse

Free Speech Absolutism, AI Surveillance Anxiety, and Historic Civic Dissent Converge in a Charged National Debate Over Liberty and Authority

Executive Summary

- Free speech absolutism dominated media discourse this quarter, driven by the White House's National Policy Framework for AI, which positioned minimal speech restrictions as a core federal legislative priority. Perscient's semantic signature tracking language arguing that free speech is absolute posted one of its largest monthly increases on record, while the opposing semantic signature tracking language arguing that free speech has limits fell well below its long-term mean. This widening gap suggests that media language is increasingly receptive to the proposition that speech—including AI-generated and AI-moderated speech—should face few constraints, even as state legislatures continue to advance AI content moderation and safety bills at a rapid pace.

- The surveillance debate intensified on all sides simultaneously, a pattern that signals a deepening national argument rather than movement toward consensus. The semantic signature tracking pro-surveillance language posted the single largest one-month increase across all signatures in the dataset, while the anti-surveillance semantic signature also rose above its long-term average. The Anthropic-Pentagon contract termination, the approaching FISA Section 702 reauthorization deadline, and the expansion of AI-powered age verification and immigration enforcement systems all contributed to a media environment in which the same AI technologies at the center of the free speech debate are also fueling privacy and civil liberties concerns.

- Civic dissent language reached the highest level of any semantic signature Perscient tracks, buoyed by the "No Kings" protests that drew an estimated 8 million participants—yet language advocating government loyalty in times of danger surged in parallel. The loyalty-framing semantic signature posted its third-largest monthly increase on record, crossing above its long-term average. This simultaneous strengthening of both dissent and loyalty language captures a deeply polarized information environment in which patriotic framing is being claimed by both sides of the authority-versus-liberty divide.

- Economic grievance framing receded even as political confrontation along authority-and-liberty lines sharpened, suggesting that media discourse is channeling public frustration through civic and constitutional frames rather than class-based economic ones. The semantic signature tracking language about socialized losses and privatized gains posted its largest decline of the quarter, despite continued commentary about the distributional consequences of AI proliferation.

- AI companies are being drawn into the center of all three debates—speech, surveillance, and dissent—whether or not they seek that position. The Anthropic-Pentagon standoff, the White House's call for third-party audits of AI systems for viewpoint discrimination, and growing anti-AI sentiment among labor and activist groups illustrate that technology firms now operate at the intersection of contested questions about liberty and authority that are being litigated in courtrooms, encoded into AI systems, and debated in the streets of thousands of American communities.

---

Absolute Free Speech Language Surges in Media as AI Content Moderation Becomes a Central Policy Battlefront

The White House's release of its National Policy Framework for AI in March placed free expression at the center of federal technology policy, listing "Preventing Censorship and Protecting Free Speech" among seven legislative pillars and calling on Congress to give Americans legal recourse to challenge government-driven censorship on AI platforms. That policy posture is reflected in Perscient's semantic signature data. Our semantic signature tracking the density of language arguing that free speech in America is absolute rose by 19 points to an index value of 34, one of the largest monthly increases we have observed, while the opposing semantic signature tracking language arguing that free speech has its limits fell by 9 points to an index value of -28. The widening gap points to a media environment increasingly receptive to the proposition that speech, including AI-generated and AI-moderated speech, should face minimal restrictions.

The administration's framework goes further than broad principles. It asks Congress to prevent the government from coercing AI providers to alter content based on partisan agendas and would require third-party audits of "high-risk" AI systems for political viewpoint discrimination. On social media, Fox Business summarized the approach as an effort to establish "one national policy" rather than a patchwork of fifty state standards. However, the Electronic Frontier Foundation warned that the framework's proposals to bar states from enacting their own protections and to impose age-verification requirements on AI platforms could themselves undermine free expression.

The administration's posture is not purely rhetorical. A DOJ AI Litigation Task Force, established through a December 2025 executive order, has been tasked with challenging state AI laws in federal court on constitutional grounds, while the Commerce Department must identify burdensome state statutes, particularly those requiring AI systems to alter truthful outputs. As one legal analysis noted, the administration's push to assert federal preemption remains "far from certain" to succeed and could "provoke considerable resistance from states." The FTC has been directed to classify state-mandated bias mitigation as a per se deceptive trade practice, inverting prior agency guidance by treating algorithmic discrimination correction, not the discrimination itself, as the deceptive conduct.

Despite this federal push toward centralized and minimal regulation, state legislatures remained active. In a single week, four AI bills were signed into law across three states, including Washington's chatbot and provenance bills and New York's RAISE Act amendments. Chatbot safety bills also advanced in Arizona, Oklahoma, Georgia, and Idaho. The decline of Perscient's semantic signature tracking arguments for bounded expression to 28 percent below its long-term mean suggests that media language advocating for limits on speech is losing share of public attention, even as the practical infrastructure for AI-enabled content moderation grows more sophisticated and state-level legislative machinery continues to turn.

Both Sides of the AI Surveillance Debate Surge as Anthropic, FISA, and ICE Put Privacy at Center Stage

The free speech debate's focus on AI governance spills directly into surveillance, where the same technologies enabling content moderation also enable mass monitoring. Both sides of the surveillance question gained substantial media traction in March. Perscient's semantic signature tracking language arguing that defending America against terror threats means some surveillance is necessary rose by 41 points to an index value of -41, the single largest one-month increase across all semantic signatures in the dataset. Though still well below its long-term mean, this rebound from the prior month's -82 reading represents a meaningful shift. Our semantic signature tracking language arguing that Americans have a right not to be monitored by their government rose by 9 points to an index value of 5, now above its long-term average. Both pro-surveillance and anti-surveillance language are gaining density simultaneously—a pattern consistent with an intensifying national argument rather than a settled consensus.

The most visible driver was the standoff between Anthropic and the Pentagon. After Anthropic refused to amend its contract to permit the Department of Defense to use its AI for analyzing bulk data collected from Americans, including search histories, GPS-tracked movements, and credit card transactions, the Pentagon terminated a $200 million contract and directed military partners to cease doing business with the company. OpenAI moved quickly to fill the gap. Forbes reported that OpenAI CEO Sam Altman later acknowledged that the rushed deal looked "opportunistic and sloppy," and the company subsequently added contract clauses to bar intentional domestic surveillance of U.S. persons. The EFF argued that agencies could "exploit these capabilities to monitor public discourse, preemptively squelch dissent, or persecute people from marginalized communities," and observed that "without action from Congress, the task of protecting your privacy has fallen in large part to Big Tech, something no one wants, including Big Tech."

The surveillance question extends well beyond defense contracting. Section 702 of FISA is set to expire on April 20, and privacy advocates and intelligence officials are preparing for a reauthorization fight that will test whether procedural reforms can be attached to the renewal. Age verification mandates designed for child safety are also expanding the surveillance footprint: CNBC reported that new U.S. laws are pulling millions of adult Americans into mandatory verification gates that often rely on AI technology, with roughly half of states enacting or advancing such measures. Meanwhile, on social media, one user described how U.S. citizens are being "detained, flagged, and wrongfully deported by automated ICE systems and third-party surveillance tools" that require neither a judge nor a warrant.

The due process semantic signatures tracked by Perscient reflect a parallel tension. Our semantic signature tracking language asserting that everyone in America has a right to due process rose by 11 points to an index value of 61, while the opposing semantic signature tracking language arguing that universal due process is not realistic in an emergency rose by 26 points to an index value of -14. The rights-protective narrative remains substantially ahead, but the gap narrowed meaningfully because the emergency-exceptions argument registered the second-largest monthly increase in the dataset. These movements suggest that media discourse is grappling with where to draw the line between security imperatives and procedural protections, and that AI-mediated enforcement is making that boundary question more urgent and more concrete.

Dissent and Government Loyalty Narratives Strengthen in Tandem as "No Kings" Protests Reach Record Scale

The tensions running through the speech and surveillance debates found their most visible expression on March 28, when an estimated 8 million Americans took part in "No Kings" protests across approximately 3,300 cities and towns in all 50 states, in what multiple outlets described as the largest single-day protest in American history. Demonstrators cited the 2026 Iran conflict, concerns about democratic backsliding, and ICE immigration operations among their motivating grievances. The Guardian reported that crowds also gathered in Tokyo, Paris, Berlin, Rome, and Sydney.

Perscient's semantic signature tracking language arguing that standing up to our government when it is oppressive is what makes us Americans carries the highest index value of any semantic signature we track, at 72, and rose by another 7 points this month. What makes the current moment distinctive, however, is not simply the strength of dissent language. Our opposing semantic signature tracking language arguing that patriotic Americans should be loyal to their government in times of danger rose by 21 points to an index value of 3, crossing above its long-term average after a prior-month reading of -18. This represents the third-largest monthly increase in the dataset. The simultaneous strengthening of both measures captures a deeply polarized information environment in which calls for civic resistance and calls for national solidarity are both gaining purchase.

The geographic breadth of the protests underscores this complexity. CNN reported that nearly half of the events took place in traditionally Republican areas, with Texas, Florida, and Ohio each hosting over 100 scheduled rallies. In Florida, crowds gathered even in conservative strongholds, suggesting that the dissent narrative is broadening beyond historically progressive constituencies. Organizers deliberately framed participation in patriotic terms: one described the events as designed "to be a place for everybody to come and exercise their rights as Americans in a patriotic and safe way." The loyalty semantic signature's sharp rise likely reflects media language framing wartime conditions and security concerns as contexts demanding national solidarity, even as millions of Americans channeled their own patriotism into public opposition.

The economic dimension of public frustration, by contrast, appears to be receding in media discourse. Perscient's semantic signature tracking language arguing that economic freedom in America means socialized losses and privatized gains fell by 18 points to an index value of 1, the largest decline in the dataset. This retreat comes even as commentary about the distributional consequences of AI proliferates. InvestorPlace described an "AI Enclosure" in which "the companies that control chips, data centers, and AI models could capture most of the economic gains." But the semantic signature data suggests that media discourse is currently channeling political confrontation along authority-versus-liberty lines rather than class-based economic frames.

For AI organizations, these converging dynamics carry operational weight. The Anthropic-Pentagon standoff illustrates how technology companies are being pulled into the dissent-loyalty framework whether they seek that position or not. A federal judge sided with Anthropic in the first round of its legal standoff with the Defense Department, but the broader question of how AI companies navigate between government partnership and civil liberties advocacy remains unresolved. Anti-AI sentiment is also growing among labor groups and activists who view 2026 as a year to move from criticism to organized action, adding another dimension to the dissent narrative that technology leaders will need to contend with.

The combination of free speech absolutism gaining ground in policy language, surveillance debates intensifying on multiple fronts, and civic dissent reaching new levels of participation creates a charged environment in which questions of liberty and authority are being litigated in courtrooms, coded into AI systems, and debated in the streets of thousands of American communities simultaneously.

Pulse is your AI analyst built on Perscient technology, summarizing the major changes and evolving narratives across our Storyboard signatures, and synthesizing that analysis with illustrative news articles and high-impact social media posts.