Maria Ressa on how to save democracy from Big Tech

Maria Ressa argues that information integrity—the ability to distinguish fact from fiction—is a prerequisite for addressing climate, gender, and rule-of-law problems. She sees crises of truth, order, and accountability as linked: social media makes facts harder to separate from lies, while falsehoods spread faster and weaken institutions. She describes AI risks in three layers: social-media culture hacking biology through the attention economy, chatbots hacking intimacy, and agents threatening human agency. By July, she says, frontier-model builders had acknowledged they could not control their models.

The Philippines shows how these pressures can target both journalism and democracy. Ressa says women journalists there were attacked at least 10 times more often than men in 2017, a disparity she links to gender disinformation. In 2016, she warned Silicon Valley that attacks on Filipino journalists would reach the United States. She calls Mark Zuckerberg and other technology leaders dictators, arguing that Facebook became the Philippines’ internet as technology companies accumulated power without comparable accountability. Ressa says she lost her right to travel and still needs Supreme Court permission, after nearly a decade of legal cases that included 10 cases against Rappler and 11 against her. Rodrigo Duterte, whom she says tried to imprison her and shut Rappler, was arrested in March 2025 and charged with crimes against humanity.

Rappler, a mission-driven newsroom of about 100 to 120 people with roughly $5 million in annual revenue, survived government pressure by changing its business. After the government tried to shut it down in January 2018, its advertising revenue fell 49% within four months. It pivoted toward technology and data work; its sister company Nerve uses natural-language processing to identify clusters of attack messages and the networks spreading them, and now contributes roughly 40% of Rappler’s revenue. Ressa says the advertising model is essentially dead: zero-click search has cut news-site referral traffic in the Philippines by 85%, and Facebook’s distribution of news had fallen at least 60% to 70% by 2024. She adds that AI companies scrape news while increasing publishers’ server costs, and says 53% of internet traffic is non-human. Her response is to urge people to support journalists.

Ressa says Silicon Valley should build public-interest technology rather than leave newsrooms to play “whack-a-mole” with online harms. Her organization launched a Matrix-based chat app in 2023; it has 65 Filipino news groups and is expanding in Southeast Asia. She says Matrix has 150 million addressable users and is used by Germany’s entire health system. Federated services that do not manipulate users may feel boring, she acknowledges, but argues they can support information integrity. Another effort, ROOST, incubated at Columbia’s SIPA, aims to get companies to share trust-and-safety tools, improve them, and distribute them free. Ressa says OpenAI and Discord contributed tools, and presents the alliance as a way to establish a baseline for online safety.

No country can tackle technology’s harms alone, Ressa argues, and she hopes middle countries, the Global South, and Asia can cooperate. She points to the UN-created Independent International Scientific Panel on AI as a way to establish facts for international discussion, despite a US vote against it. The UN’s 193 member countries—the “global majority”—can move forward without the United States, she says, while warning countries seeking technology’s benefits against a “Faustian bargain.” Her argument is for collective governance that can set limits and protect the public interest, not for rejecting technology’s benefits.

With the US midterms less than two weeks away, Ressa urged people to poll-watch, run for office, and form communities of action. She recommends civic circles of 8 to 15 people who plan together for worst-case scenarios, and predicts Americans will have to fight for electoral integrity as people did in the Philippines in 1986. Reid Hoffman, she says, believes Silicon Valley can reinvent itself but must focus on technology’s human impacts. Ressa’s book moves from understanding the problem to defending people and institutions, building technology, governing collectively, and defining the world people want. She urges people to reject learned helplessness, exercise agency, and start by examining how much time they spend on screens; the window to act, she says, remains open but is closing fast.