Category: Investigative Journalism

A $10 Million Vote for People-First AI

The headline is crisp: the MacArthur Foundation is committing $10 million in aligned grants to the new Humanity AI effort — a philanthropic push that sits inside a much larger, $500 million coalition aiming to steer artificial intelligence toward public benefit. That money is more than a donation; it’s a signal. It says: the future of AI should be designed with people and communities in mind, not simply optimized for speed, scale, or shareholder returns.

Why this matters right now

We’re living through a rapid pivot: AI is no longer a niche research topic. It’s reshaping how people learn, how news is reported, how work gets organized, and how public decisions are made. That pace has created a glaring mismatch — powerful technologies rising faster than institutions, norms, or public understanding. Philanthropy’s new role here is pragmatic: fund research, build civic infrastructure, and support the institutions that translate technical advances into accountable public outcomes.

• The $10 million from MacArthur is aimed at organizations working on democracy, education, arts and culture, labor and the economy, and security.
• The broader Humanity AI coalition plans to direct roughly $500 million over five years, pooling resources across foundations to amplify impact and avoid duplicate efforts.

What the grants will fund (the practical pieces)

The initial MacArthur-aligned grants are deliberately diverse: universities, research centers, journalism networks, and civil-society groups. Expect funding to do things like:

• Scale investigations into AI and national security.
• Support public-interest journalism that holds AI systems and companies accountable.
• Build tools and infrastructure for civil-society groups to use and audit AI.
• Convene economists, policymakers, and labor experts to measure and prepare for AI’s workforce effects.
• Create global forums that connect social science with technical development.

These are practical investments in the civic plumbing needed to make AI responsive to human values, not just technically impressive.

The larger context: philanthropy as a counterweight

Tech companies and venture capital continue to drive the research and deployment of large-scale AI models. That private momentum brings enormous benefits — and risks: concentration of power, opaque decision-making, cultural capture of creativity, and economic dislocation. A coordinated philanthropic effort does a few things well:

• It funds independent research and watchdogs that companies and markets don’t naturally prioritize.
• It supports public-facing education and debate so citizens and policymakers can participate knowledgeably.
• It enables cross-disciplinary work (law, social science, journalism, the arts) that pure engineering teams rarely fund internally.

In short: philanthropy can nudge the ecosystem toward systems that are legible, accountable, and distributed.

Notable early recipients and what they signal

Several organizations receiving initial grants illuminate the strategy:

• AI Now Institute — resources to scale work on AI and national security.
• Brookings Institution’s AI initiative — support for policy-bridging research.
• Pulitzer Center — funding to grow an AI Accountability Network for journalism.
• Human Rights Data Analysis Group — building civil-society AI infrastructure.

These groups aren’t trying to beat companies at model-building. They’re shaping the social, legal, and civic frameworks needed to govern those models.

A few tough questions this effort faces

• Coordination vs. independence: pooled efforts can avoid duplication, but philanthropies must protect grantee independence to ensure credible critique.
• Speed vs. deliberation: AI moves fast. Can multi-year grant cycles and convenings keep pace with emergent harms?
• Global reach: many harms and benefits are transnational. How will funding balance U.S.-centric priorities with global inclusivity?
• Measuring success: outcomes like "better governance" or "safer deployment" are hard to measure, complicating evaluation.

Funding is an important lever — but it can’t substitute for good public policy and democratic oversight.

What this means for stakeholders

• For policymakers: expect richer, evidence-based briefs and cross-disciplinary coalitions pushing for clearer rules and standards.
• For journalists and civil-society groups: more resources to investigate, explain, and counter opaque AI systems.
• For educators and labor advocates: funding and research to help design equitable integration of AI into classrooms and workplaces.
• For the public: clearer communication and tools to engage in debates that will shape the rules governing AI.

How this fits into the broader timeline

This announcement is part of a wave of recent philanthropic attention to AI governance. Unlike earlier eras when foundations might have funded isolated tech projects, the Humanity AI coalition signals a coordinated, sustained investment across cultural, economic, democratic, and security domains — an acknowledgement that AI’s societal consequences are broad and interconnected.

What to watch next

• The pooled Humanity AI fund’s grant-making priorities and application processes (timelines and transparency will be important).
• Early outputs from grantees: policy proposals, investigative reporting, civic tools, and educational pilots.
• Coordination with government and international bodies working on AI norms and regulation.

Key points to remember

• MacArthur’s $10 million is strategically targeted to organizations that can shape AI governance, public understanding, and civic infrastructure.
• Humanity AI represents a larger, collaborative philanthropic push (about $500 million over five years) to make AI development more people-centered.
• The real leverage is in funding independent research, journalism, and civic tools — functions that markets alone poorly provide.
• Success will depend on speed, global inclusion, measurable outcomes, and preserving independent critique.

My take

Investing in the institutions that translate technical advances into accountable social practice is a smart, necessary move. Technology companies are incentivized to move fast; funders like MacArthur can invest in pause—space for scrutiny, public education, and inclusive policymaking. That pause isn’t anti-innovation; it’s a buffer that lets societies choose what kinds of innovation they want.

If Humanity AI and its grantees keep their focus on measurable civic outcomes and maintain independence, this could be a turning point: philanthropy helping create the norms, tools, and institutions that ensure AI augments human flourishing rather than undermines it.

Sources

• $10 Million to Advance AI By and For People – MacArthur Foundation
  https://www.macfound.org/press/press-releases/10-million-to-advance-ai-by-and-for-people

• Humanity AI Commits $500 Million to Build a People-Centered Future for AI – MacArthur Foundation
  https://www.macfound.org/press/press-releases/humanity-ai-commits-500-million-to-build-a-people-centered-future-for-ai

• Foundations want to curb AI developers' influence with $500 million aimed at centering human needs – Associated Press
  https://apnews.com/article/1038b76f0ae4ef3d94095120815a65d0

A ratings service says the FTC is trying to strangle it — and the First Amendment is now part of the fight

The headline reads like a legal thriller: a company that assigns "trust scores" to news websites has sued the Federal Trade Commission, accusing the agency of weaponizing regulatory power to cut it out of the advertising ecosystem. It's NewsGuard versus the FTC, fronted by Chairman Andrew Ferguson — and the dispute raises three big questions: who gets to police the media marketplace, when does regulation become censorship, and how much power do ad buyers and agencies hold over what counts as “acceptable” news?

Why this matters (hook)

• Advertisers funnel billions of dollars through a handful of ad agencies. If those agencies can't or won't buy inventory adjacent to particular outlets, the outlets' survival and audiences are affected.
• Independent evaluators like NewsGuard say they help brands avoid reputational risk and help readers assess reliability. Critics say these ratings can be subjective or politically skewed.
• When a regulator uses merger remedies or investigations that have the effect of freezing a ratings company out of the market, the stakes shift from commercial competition to free-speech and due-process questions.

Quick takeaways

• NewsGuard filed a lawsuit in early February 2026 alleging the FTC burdened it with sweeping document demands and inserted merger conditions that effectively bar major ad agencies from using its ratings. (Filed Feb. 6, 2026.) (washingtonpost.com)
• The contested merger remedy arose in the Omnicom–Interpublic transaction; the FTC’s order reportedly prevents those ad holding companies from basing ad buys on “journalistic standards or ethics” set by third parties — language NewsGuard says was crafted to target it. (washingtonpost.com)
• NewsGuard argues the FTC’s actions violate the First and Fourth Amendments and amount to government censorship of a private service. The FTC and some conservatives argue NewsGuard has a political slant and has inflicted commercial harm on certain outlets. (washingtonpost.com)

What NewsGuard does and why advertisers use it

NewsGuard, launched in 2018 by media veterans including Steven Brill and Gordon Crovitz, uses human journalists to score sites on nine transparency and credibility criteria and publishes a “nutrition label” explaining each score. Brands and agencies have used these ratings to reduce ad placement near sites they judge risky, and browser extensions surface those trust scores to consumers. NewsGuard emphasizes transparency in its methodology and publishes the criteria it applies. (newsguardtech.com)

Why advertisers care:

• Brand safety concerns: running ads next to fraudulent, extreme, or disinformation-filled content can cause reputational damage.
• Liability and client pressure: large advertisers increasingly demand oversight tools to demonstrate they’re avoiding harmful placements.
• Centralized buying power: big holding companies and ad agencies set de facto industry norms for what’s acceptable.

The FTC’s actions that sparked the lawsuit

According to NewsGuard’s complaint and reporting by The Washington Post, two lines of FTC activity prompted the suit:

• An extensive information demand: the FTC ordered broad disclosures of NewsGuard’s client lists, ratings deliberations, communications, and financials — an investigation NewsGuard says is so sweeping it chills its business and violates privacy and press protections. (washingtonpost.com)

• A merger condition in Omnicom–Interpublic approval: the FTC’s order included language preventing the combined agency from directing ad buys based on “adherence to journalistic standards or ethics established or set by a third party.” NewsGuard argues that language functions as a ban on companies using its ratings, effectively blacklisting the service. Newsmax and other conservative outlets publicly urged the FTC to broaden the language, which NewsGuard says revealed intent. (washingtonpost.com)

NewsGuard’s legal team frames these moves as retaliation driven by political disagreement, pointing to prior public criticism of the company by now-FTC Chair Ferguson. The company has asked a federal court to block enforcement of the merger condition and the investigative demand. (mediapost.com)

The competing narratives

• NewsGuard’s story: a neutral, transparent ratings firm is being targeted for its editorial judgments. The FTC is overreaching by using merger remedies and investigations to hobble a private business whose work touches on public discourse. That, NewsGuard says, raises free-speech and due-process problems. (newsguardtech.com)

• The FTC and critics’ story: regulators and some conservative outlets argue NewsGuard exercises editorial power that has real commercial effects and that its judgments may be politically biased. From this angle, the FTC’s scrutiny is about market power and potential exclusionary conduct — not censorship per se. Public comments from outlets like Newsmax influenced how the merger language was revised, suggesting industry players saw the remedy as relevant. (washingtonpost.com)

Both sides point to market realities: when ratings influence ad placement, they affect revenue flows. The novel legal wrinkle is whether a regulator may lawfully condition a merger or investigate a small ratings firm in a way that some regard as singling out protected speech.

Broader implications

• The case could reshape how third-party content evaluators operate in advertising markets. If agencies are barred from relying on such ratings, advertisers lose one tool for brand protection; if regulators are limited, they may be less able to police potential collusion or exclusionary tactics in ad buying.
• There’s a constitutional debate at the center: does the First Amendment protect the editorial judgments of a private ratings firm from regulatory interference? Conversely, do regulators have the authority to step in when a ratings product materially affects market competition or harms specific outlets?
• The dispute exposes how intertwined advertising, editorial judgments, and platform economics have become. A private score can effectively act like a traffic light for publishers; when government action changes who can see or use that traffic light, the ripple effects are political, commercial, and civic.

My take

This lawsuit sits at the intersection of market structure and speech. NewsGuard’s methodology is transparent and human-driven — that matters in an era of opaque algorithmic moderation — but its influence on advertisers gives its judgments real economic weight. Regulators worried about arbitrary exclusion in ad markets have a legitimate role; at the same time, wielding merger conditions or sweeping investigative powers in ways that single out a small player risks the appearance (and perhaps the reality) of viewpoint-based regulation.

The healthier path would be clearer rules and neutral standards for ad buyers and ratings services: transparent criteria (which NewsGuard publishes), robust appeals and correction processes for rated outlets, and merger remedies narrowly targeted at anticompetitive conduct rather than broad language that could be read as a blacklist. These guardrails would protect both market fairness and free expression.

Final thoughts

At stake is not only one company’s business but the architecture of trust in the information ecosystem. When ratings, advertisers, and regulators collide, the outcome will shape how audiences find reliable information and how publishers — of whatever stripe — survive. Courts will now have to weigh whether the FTC crossed a constitutional line or acted within its mandate to police markets. Either way, the case underscores that in today’s media economy, the line between commerce and speech is increasingly hard to draw.

Sources

• A company that rates news sites says the Trump administration is strangling it. The Washington Post (Feb. 6, 2026).
  https://www.washingtonpost.com/technology/2026/02/06/news-guard-ftc-censorship-lawsuit/ (washingtonpost.com)

• NewsGuard Sues FTC, Claims Omnicom/IPG Condition Unconstitutional. MediaPost (Feb. 9, 2026).
  https://www.mediapost.com/publications/article/412628/newsguard-sues-ftc-claims-omnicomipg-condition-u.html (mediapost.com)

• NewsGuard — Website Rating Process and Criteria. NewsGuard (last updated 2026).
  https://www.newsguardtech.com/ratings/rating-process-criteria/ (newsguardtech.com)

• NewsGuard Wants to Fight Fake News With Humans, Not Algorithms. Wired (Aug. 2018).
  https://www.wired.com/story/newsguard-extension-fake-news-trust-score/ (wired.com)

Epstein’s emails and the Steve Tisch revelations: why the latest document dump matters

A short, sharp scene: an email thread from 2013 shows Jeffrey Epstein offering to connect New York Giants co-owner Steve Tisch with women — one exchange even has Tisch asking, “Is she fun?” The U.S. Department of Justice’s recent release of millions of pages of Epstein-related material has forced that exchange and others back into the public eye, raising familiar questions about power, access and accountability.

This post walks through what the records show, why those details matter beyond the salacious headlines, and how to think about reputational fallout when prominent figures appear in leaked or released documents tied to criminal networks.

Why this story landed in the headlines

• The Department of Justice released a massive trove of documents related to Jeffrey Epstein and Ghislaine Maxwell in late January 2026 under the Epstein Files Transparency Act.
• Multiple news outlets reported that the files contain emails from 2013 in which Epstein repeatedly offered or arranged meetings between women and Steve Tisch, who has been a co-owner and executive of the New York Giants for decades.
• Tisch has publicly said he “had a brief association” with Epstein, exchanged some emails about “adult women,” and “did not take him up on any of his invitations” nor visited Epstein’s private island. He was not charged with any crimes related to Epstein’s trafficking.

What the newly released emails actually show

• The exchanges appear to be largely contemporaneous threads from 2013 in which Epstein proposes or confirms introductions between Tisch and various women — described by Epstein in transactional language and sometimes with details about travel, age differences, or anxieties.
• Some messages show Tisch asking pointed questions (for example, whether a woman was a “working girl” or whether she was “fun”) and responding casually when Epstein followed up about encounters.
• Other messages reference professional topics — movies, philanthropy, or invitations to sporting events — mixing conventional networking with arrangements that read as personal and sexual in nature.

(These descriptions are based on contemporaneous reporting and direct excerpts from the released files as covered by major outlets.)

A few ways to interpret these revelations

• Reputation vs. criminal liability:
  • Being named in documents or receiving introductions does not equal criminal wrongdoing. Tisch has not been charged, and he denies participation in criminal acts linked to Epstein.
  • But reputational harm can be swift and enduring for public figures tied—even peripherally—to criminal networks, particularly in sex-trafficking scandals.
• Power dynamics and plausibility:
  • The exchanges exhibit the social choreography that allowed Epstein to act as a broker of introductions between wealthy men and vulnerable or young women. That pattern matters because it helps explain how trafficking networks exploited influence and financial incentives.
• Media and institutional response:
  • Teams, leagues, studios and foundations often respond defensively or with distance when board members or executives are implicated. Statements of regret, clarification of limited contact, or policies review are typical first steps — but not always sufficient to restore public trust.

What we should ask next

• Transparency: Will institutions connected to named individuals disclose any internal reviews or conclusions about conduct and associations?
• Context and corroboration: Do the emails stand alone, or are there additional documents, witness statements or contemporaneous evidence that further clarify intent and actions?
• Policy: How will sports franchises and cultural institutions update vetting and governance to reduce the risk of leaders being entangled in abusive networks?

What to remember

• Released emails indicate that Jeffrey Epstein acted as a connector between prominent men and women; they show social introductions and suggestive exchanges involving Steve Tisch but do not prove criminal conduct by Tisch.
• The public and institutions reasonably expect clearer explanations from those named in the files — both about what happened and about steps taken since to address any ethical lapses.
• Document dumps create headlines, but the long-term consequences fall on how organizations and individuals handle accountability, transparency, and prevention.

My take

The Epstein file releases are ugly, necessary reminders of how influence and commerce can cloak predatory behavior. When powerful people show up in those documents, we shouldn’t leap straight to assumptions about criminality — but we also shouldn’t minimize the moral responsibility that comes with wealth and leadership. The right first moves are clear: full transparency from institutions, independent review where warranted, and public policy that makes it harder for exploiters to operate in plain sight. The real test is whether cultural and legal systems learn from these revelations or simply file them away as another scandal headline.

Sources

• Epstein frequently connected NY Giants co-owner Steve Tisch with women, emails show — ABC News (video and article).
  https://abcnews.go.com/US/video/epstein-frequently-connected-ny-giants-owner-women-emails-129756751

• Emails reportedly show Epstein scouted women for New York Giants co-owner Steve Tisch — The Guardian.
  https://www.theguardian.com/sport/2026/jan/30/steve-tisch-epstein-emails-giants-owner

• New York Giants co-owner Tisch says he never went to Epstein’s island and regrets associating with him — CBS News.
  https://www.cbsnews.com/newyork/news/steve-tisch-ny-giants-jeffrey-epstein-files/

• The DOJ’s document release and wider reaction — Associated Press coverage of the broader Epstein file release.
  https://apnews.com/article/682447e50bf9a3643a36c9b54ccdfa22

(Note: links above point to non-paywalled news reporting on the January 2026 release of Epstein-related documents.)