Tech News — 2026-04-18#

Story of the Day#

The US government is actively seeking access to Anthropic’s highly sophisticated ‘Mythos’ AI model to bolster its cybersecurity defenses, reflecting a frantic geopolitical race to harness AI for national security. Despite Anthropic currently battling the Trump administration in court, top officials recognize that ignoring the AI’s ability to autonomously identify and exploit zero-day vulnerabilities would be a massive strategic misstep.

Top Stories#

US Government Pushes for Access to Anthropic’s ‘Mythos’ AI · Slashdot Anthropic’s ‘Mythos’ model possesses hacking capabilities far beyond previous iterations, including the capacity to launch autonomous, end-to-end cyberattacks and act as a force-multiplier for developing biological weapons. The White House and the Office of Management and Budget are negotiating access to the tool to prepare for advanced AI-driven cybersecurity threats, despite the broader administration’s contentious relationship with the startup. This highlights a growing tension where governments must adopt potentially dangerous tech out of fear that adversaries like China might gain a decisive edge.

The RAM Shortage Could Drag On for Years · The Verge If you’re waiting for hardware prices to stabilize, brace yourself: severe DRAM shortages might stretch all the way to 2030. Heavyweight memory manufacturers like Samsung, SK Hynix, and Micron are struggling to build new fabrication plants fast enough, with meaningful capacity increases not expected online until at least 2027. Despite a 12 percent annual production hike needed to meet current demand, suppliers will likely only satisfy 60 percent of the market’s requirements by the end of 2027.

Shuttered Startups Find a New Lifeline: Selling Slack Data to AI Labs · Slashdot Failed tech startups are cashing out by selling their internal Slack messages, Jira tickets, and emails to artificial intelligence companies starved for real-world training data. Tools like SimpleClosure are actively facilitating these sales, brokering over 100 deals in the past year with payouts ranging from $10,000 to $100,000. Privacy advocates are raising alarms over the practice, as this identifiable employee data is being fed into language models without individual consent.

DOJ Refuses to Assist France’s Criminal Probe into X · Engadget The US Department of Justice has formally declined to help French authorities investigate Elon Musk’s social platform, X, for alleged algorithmic manipulation and fraudulent data extraction. The DOJ slammed the French probe in a letter, accusing it of attempting to wrongfully regulate a platform’s business activities and First Amendment rights through politically charged prosecution. French officials had previously raided X’s Paris office and summoned Musk and former CEO Linda Yaccarino for interviews.

Apple Defeats Second Attempt to Ban Smartwatch Imports · Engadget Apple has successfully dodged another import ban from the US International Trade Commission concerning its redesigned smartwatches. The ITC ruled that Apple’s tweaked blood-oxygen monitoring tech no longer infringes on patents held by medical technology company Masimo, officially terminating the case. While a major victory that keeps the wearables on shelves, Apple still owes Masimo $634 million from a separate federal jury ruling last November.

Data Center Boom Endangers the Potomac River · Engadget The Potomac River has just been named the most endangered river in the US by American Rivers, largely driven by an unprecedented surge in local data center construction. The region is slated to house roughly 1,000 data centers, straining water resources and threatening the surrounding environment alongside aging wastewater infrastructure.

Ghoulish AI: Val Kilmer Deepfake Sparks Backlash in New Western · Slashdot The first authorized generative AI performance by a major Hollywood star is facing severe online criticism. The late Val Kilmer has been digitally resurrected to portray a Catholic priest in the upcoming film As Deep As the Grave, utilizing an AI voice from Sonantic and a visual deepfake. Despite the Kilmer estate’s approval and financial compensation, critics have lambasted the trailer as a “ghoulish puppet show” that sets a highly controversial precedent for the film industry.

Judge Rules Trump Admin Violated First Amendment by Banning ICE Trackers · Engadget A federal judge has granted a preliminary injunction against the government for pressuring Apple and Facebook to remove apps and groups that tracked US Immigration and Customs Enforcement (ICE) activity. The ruling declared that the Trump administration likely violated the First Amendment by coercing platforms into suppressing protected speech.

Also Worth Knowing#

• AI Chip Startup Cerebras Files for IPO (TechCrunch): The hardware maker is going public following lucrative deals, including a reported $10 billion agreement with OpenAI and a partnership to put its chips in Amazon Web Services data centers.
• 30 WordPress Plugins Compromised After Suspicious Ownership Change (Slashdot): A malicious buyer injected SEO spam backdoors into the EssentialPlugin package, compromising sites for eight months and exposing major flaws in the plugin marketplace’s trust model.
• AWS Launches Generative AI DevOps Agent (InfoQ): Amazon Web Services has released an autonomous AI agent that investigates incidents, correlates telemetry data, and actively troubleshoots production issues without human prompting.
• Pulumi Adds Native Bun Runtime Support (InfoQ): Infrastructure-as-code developers can now execute Pulumi programs using the high-speed Bun runtime, allowing native TypeScript execution and eliminating the need for a separate Node.js installation.
• US College Students Abandon CS for Data Science and Engineering (Slashdot): Traditional computer science majors saw an 8 percent drop in enrollment last year as students splintered off into specialized fields like data analytics, AI, and mechanical engineering.