🤖 HackAI Newsletter — Early April 2026

Previous Issue

Missed our last update? Read the March 2026 newsletter →

🗓️ Upcoming Events

💻 HackAI x PostHog — Monday, Apr 20 (6pm-8pm) at Capital Factory — Dive into product analytics and learn how to leverage data for your AI projects. Hands-on workshops, networking, pizza, and great conversations. Don't miss it! RSVP here →

⚡ AI Build Sprint @ LaunchD — Saturday, Apr 25, 3:00–7:00 PM — Our friends at Launchd are running a 4-hour build sprint for young Austin AI builders. You've got 4 hours. Build anything. $200 in cash prizes + Claude Pro subscriptions for the best ships. Judges pick winners, but everyone leaves with something built. RSVP here →

🌐 Big AI Headlines

🛡️ New: Anthropic Launches Project Glasswing with AWS, Apple, Google, Microsoft, NVIDIA & More Anthropic just unveiled Project Glasswing, a new initiative bringing together AWS, Anthropic, Apple, Broadcom, Cisco, CrowdStrike, Google, JPMorganChase, the Linux Foundation, Microsoft, NVIDIA, and Palo Alto Networks to secure the world's most critical software. The reason? A new unreleased frontier model called Claude Mythos Preview that has already found thousands of high-severity vulnerabilities, including some in every major operating system and web browser. Highlights: a 27-year-old vulnerability in OpenBSD that let attackers remotely crash any machine, a 16-year-old FFmpeg bug that automated testing tools had hit five million times without catching, and a chain of Linux kernel vulnerabilities allowing full privilege escalation. Anthropic is committing up to $100M in usage credits plus $4M in donations to open-source security organizations. The catch: the same capabilities that help defenders can help attackers. 🐛 Learn more →

⚔️ Update: Anthropic Wins Preliminary Injunction Against Pentagon Last time: Trump directed federal agencies to "immediately cease" use of Anthropic tech, Defense Secretary Hegseth labeled Anthropic a "supply chain risk," and Anthropic sued on March 9. A federal judge said the Pentagon's actions "look like an attempt to cripple" the company. Big update: A federal judge in San Francisco granted Anthropic's request for a preliminary injunction in its lawsuit against the Trump administration. Judge Rita Lin's order bars the Trump administration from implementing, applying or enforcing the president's directive, and hampers the Pentagon's efforts to designate Anthropic as a threat to U.S. national security. In the ruling, Lin wrote that "punishing Anthropic for bringing public scrutiny to the government's contracting position is classic illegal First Amendment retaliation." A final verdict could still be months away — but round one goes to Anthropic. Learn more →

🧬 New: Anthropic Acquires Biotech Startup Coefficient Bio for $400M Anthropic is making its first major biotech move — a $400M acquisition of stealth AI startup Coefficient Bio, aimed at AI-driven biology applications. Paired with Glasswing above, it's a signal that Anthropic is rapidly expanding beyond general-purpose chat into specialized, high-stakes domains. Learn more →

💰 New: OpenAI Alums Launch "Zero Shot" — a $100M AI Fund A group of OpenAI OGs — including Evan Morikawa (former head of applied engineering during the launch of DALL·E and ChatGPT through Codex), Andrew Mayne (OpenAI's original prompt engineer), and Shawn Jain — just closed the first tranche of a $100 million fund called Zero Shot, a play on the AI training term. They've already backed Worktrace AI and Foundry Robotics, with a third investment still in stealth. Fun bearish take from Mayne: he thinks most iterations of vibe coding will get crushed because the model makers will quickly make subscriptions to such platforms feel unnecessary. 👀 Learn more →

⚙️ New: Cognichip Raises $60M to Let AI Design the Next Generation of AI Chips Cognichip just raised $60 million in new funding led by Seligman Ventures, with notable participation from Intel CEO Lip-Bu Tan, who will be joining Cognichip's board. The pitch: their physics-informed foundation model built specifically for semiconductor design can reduce design effort by up to 75% and accelerate timelines by roughly 50%. If it works, it breaks the circular bottleneck where AI needs better chips but better chips take years to design. Learn more →

🌍 New: Spain's Xoople Raises $130M Series B for AI Earth Mapping Xoople closed a $130M Series B to build high-resolution 3D maps of Earth for AI training data and geospatial applications. Learn more →

🤠 Texas AI Spotlight

🏭 Update: Intel Joins Musk's Austin "Terafab" Mega-Project Last time: Musk unveiled the $20B+ Advanced Technology Fab at the Seaholm Power Plant — a Tesla/SpaceX/xAI joint venture on Tesla's East Travis County campus targeting one terawatt of annual compute, with potential $40B+ total economic impact. Big update: Intel is joining the Terafab project, adding serious silicon muscle to what's already shaping up to be the largest AI chip facility announcement in Texas history. The East Austin terafab just became an even bigger deal for the region's AI infrastructure bet. Learn more →

🏛️ New: Texas Legislature Targets Data Centers & AI Regulation On April 3, the Texas Lege outlined interim charges prioritizing data center impacts on property rights and community welfare for the 2027 session. Studies will cover AI in healthcare, fraud detection, and the legal sector — plus regulations flowing from HB 149/SB 1964. Learn more →

💸 Update: Texas Set to Become World's #1 Data Center Hub Last time: ERCOT flagged 24 gigawatts of new data center demand by 2031 — basically another Houston metro on the grid — with Chevron, Exxon, and Diamondback building gas plants for West Texas data centers. Now: JLL's 2026 Outlook projects Texas will be the largest data center market on Earth by 2030, capturing a significant share of the anticipated $3 trillion in AI-related construction investment. The state's data center capacity is expected to nearly double within the next decade. Learn more →

🏆 New: Austin AI Alliance Honors Texas State University The Austin AI Alliance gave Texas State's Center for Analytics and Data Science its Advance: Innovation & Research Award. The NSF-funded program has 40+ faculty and roughly 900–1,100 learners driving federal grants and pioneering interdisciplinary AI education. Learn more →

🗣️ This Month's AI Debate

Should frontier AI companies be building offensive cyber capabilities at all — even for defense?

Project Glasswing puts the question on the table in a way nobody can dodge. Claude Mythos Preview can autonomously find zero-days that survived decades of human review. Anthropic is sharing it with 12 launch partners and 40+ critical infrastructure orgs to defend — but the same weights, in the wrong hands, could power the most sophisticated cyberattacks in history. Last month we debated Pentagon contracts; this month the question is sharper.

🛡️ Pro Build It: Defenders need a decisive edge Attackers will get these capabilities eventually — the only question is whether defenders get them first. A 27-year-old OpenBSD bug that Mythos found in hours is the kind of thing state-sponsored actors have been hoarding for years. Democratizing this tech to maintainers and CISOs is how the good guys stay ahead.

⚠️ Con Don't Build It: You can't un-ship a weapon Once a model with these capabilities exists, it leaks — through exfiltration, insider access, or eventual open-source replication. "$100M in credits and a 90-day report" isn't a containment strategy. The responsible move is to not train the model in the first place.

💬 What do you think? Bring your hot takes to the next HackAI Meetup!

Community CTA

Got an AI project, job posting, or demo to share? Reply to this email or message us on Meetup — we'd love to feature you!

Stay curious, Austin 🤖💡 — The HackAI Team, Reid & Bryce