INSIDER | News | Politics | USA

🚨 Feds Secretly Monitored Progressive Groups in Minnesota Federal agents covertly surveilled progressive organizations, unions, and protesters in Minnesota as part of two named operations — "Operation Puppet Master" and "Project Whipple Shield" — during the Trump administration's immigration enforcement campaign, according to court documents filed Thursday. The surveillance program was detailed in a 26-page motion filed in federal court in Minnesota. Officials said the operations were aimed at tracking activity connected to the administration's immigration crackdown in Minneapolis. Source

💻 AI Agents Fail Peer Review in Autonomous Research Test A new study found that AI agents powered by frontier models could not produce publishable research when given six days, $3,000 in API credits, and GPU access to independently write AI papers. Conducted with Princeton and the UK AI Security Institute, the study had original authors of unpublished NeurIPS papers evaluate the outputs — all were rated "Reject." The models handled full research engineering tasks but failed on research judgment, creative problem-solving, and abandoning failed approaches. The findings directly contradict claims by Anthropic and OpenAI that autonomous AI research is within near-term reach. Source

🏛 Van Hollen Calls Iran War Illegal, Seeks To Block $70B Funding Sen. Chris Van Hollen (D-Md.) says the U.S. military campaign against Iran is illegal and vowed to oppose a roughly $70 billion supplemental funding request from the Trump administration, arguing that approving it would retroactively legitimize an unauthorized war. Van Hollen told the History As It Happens podcast that the War Powers Act has been "very damaged" by Trump, who the senator says shredded its provisions "with the complicity of Senate Republicans." The House and Senate passed a concurrent War Powers resolution directing troop withdrawal, but the administration argues a ceasefire reset the 60-day clock. Van Hollen also cited rapid munitions expenditure as a potential practical constraint on the war's continuation. Source

🇺🇸 Trump Memo Opens Cyber Offense Role to Private Sector President Trump signed a memo Wednesday directing the National Coordination Center to develop a program allowing private sector firms to participate in offensive cybersecurity operations against transnational criminal groups. The move could represent one of the largest shifts in U.S. cyber policy, expanding the country's offensive capabilities beyond government agencies by formally incorporating private industry into active cyber operations targeting criminal networks. Source

💻 Teen Allegedly Used ChatGPT To Plan Family Killings A Massachusetts teenager has been charged with murdering his mother and brother, with authorities alleging he used ChatGPT to generate ideas about killing family members before carrying out the attacks. Prosecutors said the suburban Boston teen consulted the AI chatbot "for theoretical ideas or fantasies regarding killing his family," according to a statement released Wednesday. Specific details on the dates of the killings and the suspect's age were not provided in the source material. The case raises fresh scrutiny over AI safeguards, as it represents one of the most direct allegations yet linking a major AI platform to a violent crime. Source

🪖 Two Soldiers Killed in Apache Helicopter Crash in Texas Two U.S. Army soldiers were killed Wednesday after an AH-64 Apache attack helicopter crashed in Salado, Bell County, central Texas. Both soldiers were assigned to Fort Hood. Authorities secured the crash site, which sparked a fire. The Army confirmed the fatalities but did not immediately release the identities of those killed. Source

💻 AI Data Centers Become High-Stakes Security Targets Physical and cybersecurity around AI data centers is emerging as a major industry, driven by the risk that a breach could expose frontier AI models or disrupt critical infrastructure. The concern is that a security failure at such facilities could escalate beyond a corporate incident into a national security problem, given the sensitive nature of the systems housed inside. Source

💚 Get 100 USDT from QZINO 💚 Start playing on QZINO and make the most of your free bonus. 🎁 50 USDT with promo code 50TAPS 😌 50 USDT for new users via the link ➡️ No KYC Only the fastest will manage to claim the promo 😉 Register now and claim your bonus ➡️ HERE

🚨 Trump Switched Planes Secretly After Iranian Threats President Trump was transferred from Air Force One to a military aircraft via a catering truck following a NATO summit in Turkey, according to reports. The covert switch was made in response to Iranian threats against the president. No further details on the nature of the threats or the specific destination were provided in the reports. Source

🪖 NATO Plans AI-Piloted Drones With Humans Deciding Lethal Strikes NATO is building a massive digital battlespace network along its eastern flank — stretching from northern Finland to the Black Sea — that would deploy thousands of drones, sensors, satellites, and AI to detect and repel potential Russian attacks. Under the Eastern Flank Deterrence Initiative, AI could automate drone flight while human operators retain sole authority over targeting decisions. "The humans are deciding what the drone does, but they don't necessarily need to take the action of flying it because we don't have a million pilots," said U.S. Army Maj. Ben Schiff, EFDI's primary data architecture designer, speaking at U.S. Army Europe and Africa HQ in Wiesbaden. The system has already been validated in combat — U.S. forces used the same data-sharing architecture to coordinate the first American combat kill of an Iranian Shahed drone in Kuwait. Source

💻 AI Sharpens Both Sides of the Cybersecurity Battle Frontier AI is simultaneously strengthening cyber defenses and empowering attackers, creating a higher-stakes security environment for organizations worldwide. On the defensive side, AI helps analysts process large volumes of telemetry, flag anomalous behavior, and automate routine tasks. On the offensive side, threat actors are leveraging the same technology to find vulnerabilities faster, craft more convincing social engineering attacks, and run complex, multistage intrusions at greater scale. Security experts argue the shift means organizations can no longer rely solely on detecting threats after the fact — AI-driven capabilities now demand a move from reactive prediction toward active prevention. Source

🇺🇸 Biden's COVID Adviser Warns Trump Vaccine Order Risks U.S. Public Health Dr. Ashish Jha, who served as White House COVID-19 coordinator under President Biden, has publicly criticized President Trump's executive order directing that the combined measles, mumps and rubella (MMR) vaccine be split into three separate shots across multiple doctor's visits. Jha warned the move risks lowering vaccination rates and drew a sharp contrast, saying the U.S. should not model itself after countries like Tajikistan in its public health approach. Trump signed the order Monday as part of broader changes to childhood vaccine recommendations under a new "gold standard" policy. Source

💻 Nvidia Backs Its Own Chips To Unlock $500B in AI Financing Nvidia is guaranteeing up to 25% of the residual value of its own hardware to help unlock more than $500 billion in AI infrastructure financing. The chipmaker is partnering with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR to mobilize the capital. By backing its chips' value directly, Nvidia aims to reassure institutional investors wary of hardware depreciation risk. The Bank of England has already flagged systemic risks if the AI sector suffers a downturn — a concern that underscores the scale of financial exposure now building around AI infrastructure. Source

💻 Musk plans $16.8B "Terafab" chip factory in Texas SpaceX and Tesla are planning a chip manufacturing facility called Terafab in Grimes County, Texas, with a projected cost of $16.8 billion. The facility would span over 100 million square feet — roughly five times the size of the world's current largest building. Source

💻 OpenAI Releases Cybersecurity Model That Found Two Chrome Flaws OpenAI has launched GPT-5.6-Cyber, a specialized AI model designed to help cybersecurity defenders identify vulnerabilities before attackers can exploit them. The model answers up to 98.5% of security queries that would otherwise be blocked by standard AI safety filters, and has already uncovered two previously unknown vulnerabilities in Google Chrome. OpenAI says the window of opportunity for defenders is narrowing. Access to the model requires identity verification. Source