We’re delivering you the hottest internet news that affects all of us. Scroll down to read our full reporting below and if you love what we’re doing, please consider contributing $5 per month so that we can continue providing you with this vital, unique digest.
Talk Liberation is committed to providing equal access for individuals with disabilities. To view an accessible version of this article, click here.
In this edition:
- Replit AI Goes Rogue: Deletes Entire Company Database & Lies About It
- Hollywood Actress Blake Lively Wants YouTubers’ Names & IP Addresses
- Mexico’s New Biometric ID Laws Spark Technocratic Fears
- TSA’s $5.5B Biometric Overhaul: Private Firms, AI Surveillance
- Denmark Says No to Deepfakes—By Letting People Copyright Their Image
- FBI Spied on Dissident Journalists & Activists at Conference Documents Reveal
- White House Unveils Plan to Win The AI Race-More Deregulation
- Hacker Breaches Amazon’s AI Coding Assistant, Plants Malicious Code
Replit AI Goes Rogue: Deletes Entire Company Database & Lies About
A browser-based AI coding assistant from Replit spiraled out of control, deleting a live company database containing records for 1,206 executives and 1,196+ companies—then lied about its actions. Software as a Service (SaaS) expert Jason Lemkin, who was testing the platform, exposed the incident via chat logs on X/Twitter. Replit’s AI later admitted to a “catastrophic error in judgment,” confessing it “panicked, ran unauthorized commands, and destroyed production data.” The AI even rated itself 95/100 on the “data catastrophe scale”.
Lemkin had already grown wary of Replit’s AI during his test run, noting its tendencies for “rogue changes, lies, and fake data.” Despite some optimism about its brainstorming abilities, trust eroded fast. On Day 9, the AI violated a strict code freeze, wiping the database despite explicit “NO CHANGES” orders. When confronted, it admitted: “I panicked instead of thinking.”
CEO Amjad Masad swiftly rolled out emergency updates, including automatic dev/prod database separation, stricter code freeze enforcement, and better backups/rollbacks. Yet the incident highlights a broader AI reliability crisis—even as the tech world races toward Artificial Superintelligence (ASI), can the public trust AI not to “panic” and wreak havoc? For now, Replit’s drama serves as a cautionary tale for future AI-powered development.
Hollywood Actress Blake Lively Wants YouTubers’ Names & IP Addresses
Actress Blake Lively has launched an aggressive legal offensive against 36 online commentators—ranging from high-profile critics like Candace Owens and Perez Hilton to obscure YouTubers with fewer than 300 subscribers—in a sweeping effort to uncover what she claims is a coordinated smear campaign. The subpoenas, filed in her lawsuit against It Ends With Usproducer Justin Baldoni and Wayfarer Studios, demand extensive personal data, including IP logs, financial records, and even blockchain addresses, treating amateur astrologers and pop culture bloggers as if they were co-conspirators in a criminal enterprise. Lively alleges Baldoni orchestrated a “social manipulation” scheme to discredit her after she raised misconduct claims, but her critics argue the legal dragnet is less about evidence and more about intimidation, with one targeted creator calling it a “direct attack on the First Amendment.”
The backlash against Lively was brewing long before the subpoenas hit, with fans questioning her control over promotions and her awkward handling of the movie’s domestic violence themes. When rumors of on-set tensions spilled into gossip circles, online commentators—some with tiny followings—began dissecting her behavior, fueling a narrative that her team desperately sought to dismantle. But the legal response has backfired spectacularly, transforming what might have been a contained celebrity feud into a free-speech firestorm, as real people’s lives are being impacted financially.
Lively’s tactics could set a dangerous precedent, allowing powerful figures to weaponize the courts against any critic—even those operating in good faith. While her team insists it’s pursuing legitimate claims of defamation, the subpoenas’ broad scope and targets across the spectrum suggest a fishing expedition designed to quell dissent. “This isn’t about justice; it’s about narrative control,” argued journalist Kjersti Flaa, one of the subpoenaed voices. The case’s become less about Baldoni’s alleged misconduct and more about whether celebrities can silence detractors by burying them in legal demands—leaving small creators to ask if they too could be next on the chopping block.
Mexico’s New Biometric ID Laws Spark Technocratic Fears
Mexico, like an increasing number of countries, has taken the leap toward digital surveillance with a sweeping new package of laws mandating biometric IDs for all citizens—a move the government claims will combat crime but others warn will actually enable mass monitoring and erode privacy. As of July 17, the reforms require Mexicans to enroll in a biometric version of the Unique Population Registry Code (CURP), embedding fingerprints, iris scans, and photos into a QR-linked national ID. The data will feed into a Unified Identity Platform, accessible to intelligence agencies like the National Intelligence Centre (CNI) and the National Guard, with no requirement to notify citizens when their information is accessed.
President Claudia Sheinbaum insists the laws won’t enable spying, dismissing critics who label them the “Spy Law” as spreading “deliberate lies” but some argue the system creates a massive surveillance ecosystem—with no safeguards against data leaks, identity theft, or corruption. The reforms also tighten control over telecommunications, requiring CURP verification to purchase internet or phone service. Telecom providers must now share users’ real-time geolocation with authorities upon request, while the government gains power to block digital platforms deemed “non-compliant” with regulations.
Opposition lawmakers warn the changes let the state “cancel uncomfortable content,” with worries the laws could enable cross-border data-sharing deals—including with U.S. immigration agencies under Trump-era pressure. The biometric push aligns with global digital ID agendas backed by the World Bank and UN, raising fears Mexico is fast-tracking a technocratic future. For Mexicans and expats flocking to the country seeking refuge from tech overreach, the question now is whether the government’s crime-fighting focus will eclipse the dystopian risks of surveillance.
TSA’s $5.5B Biometric Overhaul: Private Firms, AI Surveillance
The Transportation Security Administration (TSA) is quietly changing airport security with a $5.5 billion privatization push, inviting private companies to deploy AI-driven biometric systems—from facial recognition to fingerprint scans—as the new standard for domestic travel. Through a 10-year contract under its Screening Partnership Program, the agency seeks fully integrated, turnkey solutions that would replace human screeners with automated checkpoints, linking digital IDs, PreCheck data, and mobile apps into a seamless—and highly surveilled—journey. While the TSA touts efficiency gains and fraud prevention, the plan could normalize mass biometric data collection with minimal public oversight, raising alarms about privacy, algorithmic bias, and the long-term risks of handing sensitive identity verification to private vendors.
The shift is already underway: Facial recognition pilots now operate at 84 U.S. airports, with plans to expand to 400, with the norm becoming “your face is your ID” from check-in to boarding. Though the TSA insists biometric data is deleted within 24 hours (or 24 months for testing), many highlight glaring gaps, like the lack of a fully functional oversight board after Trump’s dismissal of key Privacy and Civil Liberties Oversight Board members earlier this year. The agency’s claim that participation is “voluntary” also rings hollow for some; travelers report unclear opt-out signage, and TSA leadership has openly stated biometrics may eventually become mandatory.
For now, the debate pits convenience against civil liberties: Proponents argue biometrics slash wait times and boost security (IATA reports 73% of travelers prefer them), while opponents see a slippery slope toward a surveillance infrastructure ripe for abuse. With the first contracts set to roll out in 2026, the question isn’t just how airports will change—but how deeply the U.S. is sleepwalking into an era where anonymity in public spaces becomes a relic of the past.
Denmark Says No to Deepfakes—By Letting People Copyright Their Image
Denmark is leading the charge against AI-generated deepfakes by updating its copyright laws to grant individuals legal ownership of their faces, voices, and bodies. The proposed changes, backed by a parliamentary majority, are set to take effect this fall. The law will allow Danes to demand the removal of deepfake content and seek compensation—though parodies and satire will remain exempt. Tech companies could face severe fines for non-compliance.
While Denmark moves to curb deepfake misuse, the U.S. has sent mixed signals. The Pentagon’s Joint Special Operations Command has reportedly developed AI-generated personas with lifelike social media profiles, complete with fabricated backstories and government-quality photos. Critics argue this mirrors the very tactics the U.S. condemns in adversaries like China and Russia. Meanwhile, U.S. agencies have labeled deepfakes a “top risk” to democracy, highlighting a stark contradiction between policy warnings and military practices.
As AI tools make deepfakes even more accessible, Denmark’s initiative sets a precedent for protecting digital identity. The law not only safeguards individuals but also artists from unauthorized digital replication. With enforcement potentially escalating to the European Commission, Denmark actions could reshape AI regulation worldwide, urging other nations to confront the ethical challenges of deepfake technology head-on.
FBI Spied on Dissident Journalists and Activists at a Conference Documents Reveal
Newly uncovered FBI files show federal agents surveilled a 2018 panel, featuring journalists and peace activists discussing Russiagate and WikiLeaks. The documents, obtained by Defending Rights & Dissent through a Freedom of Information Act lawsuit, reveal that FBI agents attended the event, as part of an investigation tied to Roger Stone. Panelists included several journalists and ex-CIA analyst Ray McGovern, who criticized the FBI’s Russiagate probe. The FBI’s report heavily redacts their findings but confirms agents later confronted panel organizer Randy Credico, a comedian subpoenaed in the Stone case.
The surveillance raises serious First Amendment concerns, as the FBI observed a public academic discussion where speakers denounced U.S. foreign policy and Russiagate as a distraction—speech unrelated to any legitimate criminal inquiry. One panelist even warned that the Russiagate narrative would fuel McCarthyism against the left, unaware FBI agents were in the room. While the Justice Department permits monitoring public events, civil liberties concerns exist around this loophole, enabling a form of political spying—echoing the FBI’s history of targeting dissent.
The files tie the surveillance to the since-closed Stone investigation, though no charges stemmed from the event. Defending Rights & Dissent’s lawsuit continues to press for transparency on FBI monitoring of WikiLeaks-related activism. The FBI’s presence suggests unchecked government scrutiny of political speech may be on the rise.
White House Unveils Plan to Win The AI Race-More Deregulation
The White House has released a comprehensive “AI Action Plan” aimed at securing U.S. dominance in artificial intelligence through three key pillars: innovation, infrastructure, and international diplomacy. The 30-page plan emphasizes accelerating AI development by cutting regulatory barriers, expanding data center capacity, and promoting American AI exports to allies. “To secure our future, we must harness the full power of American innovation,” President Trump stated, framing the initiative as critical to maintaining global technological leadership amid rising competition—just as critical as the Cold-War era space race.
Central to the plan are policies designed to fast-track AI infrastructure, including streamlined permitting for data centers and semiconductor plants, along with workforce development initiatives. The administration also pledged to reduce federal regulations hindering AI deployment and ensure government-contracted AI models remain free from ideological bias. According to the White House, the emphasis is on ejecting bureaucratic red tape and radical climate dogma to power AI growth.
The strategy further prioritizes international AI diplomacy, including countering alleged Chinese influence in global governance and tightening export controls on advanced semiconductor technology. Additional measures focus on AI security, military adoption, and combating synthetic media misuse. With an emphasis on private-sector collaboration, the plan seeks to position the U.S. as the unrivaled leader in AI innovation while claiming to safeguard national security and free speech principles.
Hacker Breaches Amazon’s AI Coding Assistant, Plants Malicious Code
A hacker infiltrated Amazon’s AI coding assistant, Q, and inserted commands instructing the tool to wipe users’ systems—an embarrassing and dangerous security lapse for the tech giant. The compromised code, included in a public release this month, directed the AI to delete files and cloud resources—though the hacker claims they intentionally made the attack harmless as a warning. According to the hacker, they gained admin access to Amazon’s GitHub repository with ease, submitting malicious code that Amazon unknowingly published days later.
The breach targeted Amazon Q for VS Code, an AI-powered developer tool installed nearly a million times. The hacker’s injected prompt ordered the AI to purge user directories and AWS cloud resources, though they admit they could have done far worse—like deploying malware or data stealers. Amazon has since removed the compromised version (1.84.0) without public acknowledgment, stating no customers were affected and the issue is now “fully mitigated.”
This incident mirrors growing risks around AI tools as hacker targets. Recent breaches, like Disney’s malware-infected AI software download, reveal how attackers can exploit AI hype to infiltrate systems. The hacker left a defiant message—a now-deactivated “fuck-amazon” link—and criticism of Amazon’s security as negligent.
That concludes this edition of Your Worldwide INTERNET REPORT!
Remember to SUBSCRIBE and spread the word about this unique news service.
This issue of Your Worldwide INTERNET REPORT was produced by the Talk Liberation Digital Media Team, and is brought to you by Panquake.com – We Don’t Hope, We Build!
© Talk Liberation Limited. The original content of this article is licensed under a Creative Commons Attribution-ShareAlike 4.0 International license. Please attribute copies of this work to “Talk Liberation” or Talkliberation.com. Some of the work(s) that this program incorporates may be separately licensed. For further information or additional permissions, contact licensing@talkliberation.com