India's DEPA Framework to Boost GDP by $500 Billion

Key Takeaways

• India's DEPA framework aims to add $500 billion to GDP by 2030 while enabling AI training on sensitive data using differential privacy and secure data environments
• Dulo raised $130 million in seed funding for hardware design AI, backed by Eric Schmidt, John Doerr and Marc Benioff
• Intel's data center business grew nearly 80% year-over-year to $7.3 billion in Q3 2026, driven by AI server CPU demand
• CopperPilot raised $390,000 for its Epsilon AI agent that creates circuit board designs from plain-English descriptions in hours
• Four major AI companies agreed to honor website opt-out signals through robots.txt, but already-scraped data remains a concern
• JPMorgan raised Meta's price target to $940, noting Muse AI assistant has over 10 million users and capital spending of $135-$145 billion this year
• Apple's John Ternus emphasized that AI features must reduce effort while preserving user control and human creativity
• Factory's Droid 1.0 coding agent scores 65.2% on Terminal-Bench 4 using Claude Opus 5.5 at lower cost than competitors
• India's DEPA framework uses a model-travels-to-data approach, compared in significance to UPI's 2016 launch
• Intel expects supply constraints for server CPUs to remain severe through the first half of 2027

India launches DEPA framework to boost AI training and protect data privacy

India's Ministry of Electronics and Information Technology released the DEPA for AI Training framework. It aims to let companies and startups train AI models on data from hospitals and banks while keeping raw data private. The framework could add $500 billion to India's GDP by 2030. It uses four pillars including differential privacy and secure data environments. Pilots are planned for finance and healthcare sectors.

MeitY launches DEPA handbook for secure AI data sharing

The Union Ministry of Electronics and Information Technology launched the DEPA for AI Training Handbook in Bengaluru. Secretary S. Krishnan unveiled the handbook with support from the iSPIRT Foundation. The framework helps institutions share data for AI training while keeping control of their information. It uses contractual permissions and privacy-enhancing safeguards. The system also tracks data origin and usage for transparency.

Government rolls out data-sharing framework for AI training

MeitY convened a roundtable to launch the DEPA for AI Training Handbook. The framework addresses concerns about privacy, confidentiality and intellectual property when sharing data. It allows institutions to collaborate on AI training without giving up control of their datasets. Principal Scientific Adviser Ajay Kumar Sood said the model travels to the data rather than the data traveling to the model. The framework is compared in importance to the launch of UPI in 2016.

Dulo raises $130 million seed funding for hardware design AI

San Francisco startup Dulo raised $130 million in seed funding to build AI models for hardware design. Co-founders Adrian Aoun and Sebastian Thrun started the company in 2026. Investors include Eric Schmidt, John Doerr and Marc Benioff. The funding round also included Construct Capital and Lowercarbon Capital. Dulo has not yet announced what types of hardware it will target or when products will be available.

CopperPilot raises $390,000 for AI hardware design agent

CopperPilot raised $390,000 to develop its AI design agent called Epsilon for electronics hardware. The agent takes plain-English descriptions and creates circuit board designs in hours instead of weeks. It works with tools like KiCad and Altium. The startup reports 6,000 design tasks completed and 25-fold user growth since August 2026. Founded in October 2025, CopperPilot has customers across the US, UK, Europe and India.

Intel shifts CPU production to meet surging AI demand

Intel is shifting manufacturing capacity from consumer chips to server CPUs due to unprecedented AI demand. The company's data center business earned $7.3 billion in Q3 2026, growing nearly 80% year-over-year. Intel plans to expand production of Xeon 6 processors at facilities in Ireland, Oregon, and Arizona. The company expects capital spending to reach about $20 billion in 2026 and grow significantly in 2027. Supply constraints are expected to remain severe through the first half of 2027.

AI tools on construction sites may not reduce injuries

AI software is being used on construction sites to cut paperwork and flag safety hazards. Company FYLD says customers reported up to 48% fewer serious injuries, but researchers say evidence that AI prevents injuries remains limited. Experts warn that false alarms can cause workers to ignore warnings and overreliance on AI can create false security. Researchers note that construction sites change constantly, making it hard for cameras to capture every hazard. Studies so far have mainly focused on productivity gains rather than injury prevention.

AI giants agree to website opt-outs but data is already scraped

Four major AI companies have agreed to honor website opt-out signals through robots.txt. However, experts say this does not solve the larger problem of data that has already been scraped from the internet. The agreement allows websites to block AI companies from collecting their content. Still, the concern remains that massive amounts of data were already taken without permission before these rules were put in place.

Creator builds AI Tamagotchi using Meta hardware and Claude

A reporter built a small AI pendant device called Muse Charm using a $50 Waveshare computer and Meta's software. He used Claude AI code to reprogram the device and add features like custom avatars and a Tamagotchi-style pet game. The device can take photos through a phone camera and connect to smart home gadgets. The creator said reprogramming the hardware with Claude was quick and easy. Meta has its own version of this kind of device but the reporter already had a working alternative.

JPMorgan raises Meta price target to $940 with strong outlook

JPMorgan raised its Meta price target to $940 and kept an Overweight rating despite Meta stock falling over 12% in five trading days. The analyst cited strong Q3 earnings and growth in AI advertising tools. Meta's personal AI assistant Muse has surpassed 10 million users since launching in the US and Canada. Meta's capital spending this year is expected to reach $135 billion to $145 billion for AI infrastructure. The stock faces support near $525 and resistance near $800.

Factory Droid 1.0 brings model-independent coding agent with Script mode

Factory released Droid 1.0, a coding agent harness that works with any supported model. The new version makes programmatic tool calling the default, where the model writes a script that runs tools in a sandbox. Factory reports Droid scores 65.2% on Terminal-Bench 4 with Claude Opus 5.5 at lower cost than competitors. The base prompt shrank 37%, and first paint latency dropped from 1,839 ms to 488 ms.

John Ternus centers Apple AI strategy on useful products

John Ternus emphasized product usefulness as the core of Apple's AI strategy in an interview. He argued that an AI feature must remain reliable when used with real documents, preferences, and unfinished projects. Ternus stressed that human creativity still matters, since someone must decide what the finished work should accomplish. He encouraged customers to evaluate Apple's AI by whether it reduces effort while preserving user control.

Sources

• India's DEPA Framework: A Game-Changer for AI Innovation and Data Privacy
• MeitY launches DEPA handbook to enable secure data sharing for AI training
• Govt rolls out data-sharing framework to boost AI training while protecting privacy
• Dulo Raises $130 Million in Seed Funding for Hardware Design AI
• CopperPilot, an AI design agent startup for electronics hardware, raises $390,000 in funding
• Intel shifts production capacity from consumer to server CPUs in face of 'unprecedented' AI demand — company navigating 'one of the most significant supply shortages in history'
• AI is coming to construction sites. But does it make workers safer?
• The 4 AI Giants Just Agreed to Let Websites Opt Out. They've Already Taken the Data
• I built an AI Tamagotchi before Meta could
• META Stock Price Forecast: JPMorgan Raises Price Target to $940, Sees Over 50% Upside
• Meet Factory Droid 1.0: A Model-Independent Coding Agent Harness Built on Script Mode
• John Ternus Puts Products at the Center of Apple’s AI Strategy