Agility Goes Public at $465M as Qualcomm Eyes $15B Data Center Shift

Agility Robotics Goes Public via $465 Million SPAC Merger

Humanoid robot manufacturer Agility is going public through a special purpose acquisition company (SPAC) reverse-merger that will raise $465 million in new capital from retail investors and private funding, according to TechCrunch. Spun out of Oregon State University in 2015, Agility has deployed its bipedal Digit robots across industrial facilities with nine customers, including Schaeffler, GXO, Toyota Motor Manufacturing Canada, and Mercado Libre, moving more than 100,000 heavy totes at a GXO logistics facility.

In the first six months of 2026, Agility generated $174,000 in revenue—mostly from related parties—while losing $362 million. However, CEO Peggy Johnson noted that Agility has secured $300 million in planned contracts from 30 customers for its next-generation Digit 5 robot, which is designed to operate safely alongside human workers outside isolated work cells and will be demonstrated at CES 2027 before wide availability at the end of 2027.

Agility pitches that a robot can save companies $100,000 annually under a robot-as-a-service model, with prospective customers paying $500,000 for an on-site proof of concept (four customers had signed up as of May 2026). While Digit places heavy loads such as 50-pound objects with 98% accuracy and relies on 65,000 hours of real operational data, cofounder and CTO Johnathan Hurst outlined long-term goals scaling toward 10,000 robots annually at its production facility and eventually 100 million robots.

Discovered Materials Launches $1 Million Materials Science AI Challenge

San Francisco startup Discovered Materials, a member of Y Combinator's spring 2026 batch that uses AI models to identify new substances for advanced semiconductor chips, announced a $1 million prize pool for AI models to solve five open problems in materials science, TechCrunch reported.

Co-founder Advaith Sridhar and an advisory group of university academics—including Stanford professor Felipe Jornada—defined the five challenges:

• Developing a theory to explain cuprate superconductivity at ultra-low temperatures.
• Creating a numerical method to predict the structures of interfaces between silicon and silicon dioxide.
• Formulating a numerical method to map grain structure to thermal properties for heat dissipation materials.
• Establishing a theory to predict why shining light on materials changes their behavior around superconductivity.
• Designing a numerical method to improve Density Functional Theory for modeling atomic and molecular structures.

Qualcomm Forecasts $15 Billion in Data Center Revenue by 2029

As smartphone sales declined by around 20 percent compared to 2025, Qualcomm detailed plans during its fourth-quarter 2026 earnings call to expand its AI data center business through partnerships with Amazon and Meta, according to The Verge.

Qualcomm forecasts around $650 million in data center revenue during the first quarter of 2027, rising to $5 billion by the end of 2027 and $15 billion by 2029. Compared against Qualcomm's total 2026 revenue of $10.4 billion, the company expects data center chips to eventually account for around 30 percent of its overall business, while rival AMD expects over 80 percent of its revenue to come from data centers.

OpenAI Publishes National Community Compact, Compute Blueprint, and 2028 Policy Essay

OpenAI published its National Community Compact and Compute Campus Policy Blueprint, establishing commitments for every future compute campus the company owns and operates. Building on its January Stargate Community commitments, OpenAI pledged to pay the full cost of its electricity and grid infrastructure, prioritize closed-loop or low-water cooling, publish annual site-level water and electricity usage and independent noise studies, avoid non-disclosure agreements with public officials, and invest in local priorities—citing a $4.8 million grant in Effingham County, Georgia, providing free school meals for all 14,000 students, a joint $6 million childcare commitment with SB Energy in Pike County, Ohio, and an agreement with Georgia Power allowing up to 1,000 megawatts of power curtailment during peak demand.

Separately, OpenAI chief global affairs officer Chris Lehane published an essay comparing AI's economic potential to President Franklin D. Roosevelt's 1930s New Deal and outlining five policy priorities ahead of the 2028 presidential election, TechCrunch reported. Lehane proposed tax incentives for worker AI training, education initiatives, data center expansion, and enforceable AI safety standards—drawing criticism from Kate Brennan, AI and corporate power policy lead at the AI Now Institute, who noted OpenAI's prior opposition to California's 2024 AI safety bill SB 1047 and argued that Lehane's proposals favor private corporate adoption over binding worker protections.

AI PAC Spending Results in the 2026 Midterms

Political action committees on both sides of the AI regulation debate saw mixed outcomes in Tuesday's midterm elections, according to TechCrunch. Pro-industry groups American Mission—the Republican arm of Leading the Future (LTF), backed by OpenAI president Greg Brockman and Andreessen Horowitz—and Building the Future collectively spent $4.8 million backing South Carolina Senate candidate Darline Graham, who lost, while American Mission spent more than $500,000 apiece supporting Republican House candidates Aaron Flint in Montana and Laurie Buckhout in North Carolina, both of whom also lost.

LTF's Democratic arm, Think Big, backed Florida Rep. Jared Moskowitz in his winning reelection bid after LTF spent $27.38 million in the primaries, including $8 million opposing New York Assembly member Alex Bores in June. Meanwhile, Anthropic-backed Public First contributed $3 million to a PAC supporting California State Sen. Scott Wiener, and the Guardrails Alliance—whose political director Charlie Blaettler said its goal is to make AI money politically toxic—spent more than $1 million apiece opposing Montana's Aaron Flint and California Rep. Jimmy Gomez (both of whom lost) plus $300,000 backing Gomez's victorious challenger, Angela Gonzales-Torres.

Enterprise & Consumer AI Releases: Images 2.6, DevRev, Figma, Nothing, and Uber

Across product and enterprise earnings announcements:

• OpenAI Images 2.6: OpenAI launched Images 2.6 globally in ChatGPT, Codex, and the API (via GPT-Image-2.6 Flare, which generates low-quality images in as little as six seconds, and GPT-Image-2.6 Sunburst), delivering 50% faster image generation, GIF animation, SVG exports, and a new complexity parameter as users generate 4 billion images per week (more than half a billion per day).
• DevRev Autonomous Self-Service: Enterprise platform DevRev integrated OpenAI's GPT-6.1 Sol and GPT-6 Astra into its Computer and Operator tools to execute traceable, permissioned, and reversible customer and employee actions evaluated on its Enterprise-Bench benchmark, while CTO Ahmed Bashir noted that DevRev's entire engineering team migrated to Codex and dropped its Cursor licenses, according to OpenAI News.
• Figma & Nothing: Design software company Figma announced it is allowing clients to use more AI features for the same price to prioritize long-term growth, while Nothing CEO Carl Pei announced on X that Nothing's first "AI-native product"—a productivity tool with no new screen that could not have been an app—will ship in December alongside planned AI-linked smart glasses and earbuds.
• Uber AI Agent Volume: During Uber's third-quarter earnings call, CFO Balaji Krishnamurthy and CEO Dara Khosrowshahi reported that out of around 320 million trips and orders handled during the week of Halloween in October, only four rides and 500 Uber Eats orders were booked through AI agents, TechCrunch reported. Khosrowshahi noted that booking a ride in the Uber app takes an average of 50 seconds and reordering food takes 90 seconds, though he expects the roughly eight major AI assistants currently on the market—alongside startups like Instinct ($10 billion valuation) and Meta's Muse (10 million users)—to grow to more than 10.