AI Daily Briefing — 2026-06-16

Your brain operates in the dark, yet it constantly monitors signals from your body—your racing heart, a nervous stomach, or the air filling your lungs. This inner awareness, known as interoception, processes roughly 11 million bits of sensory data every second, though only a tiny fraction reaches conscious thought. As Moriah Thomason, a neuroscientist at NYU Langone, explains, this filtering is essential for functioning. Without it, we would be overwhelmed by the flood of information our bodies generate.

Coined in 1906 by neurophysiologist Charles Sherrington, interoception remained a niche concept for most of the 20th century. But recent advances, including a 2021 Nobel Prize and new mapping tools, have sparked a surge in research. Scientists are now decoding how signals travel between body and brain, offering fresh insights into conditions like obesity, chronic pain, and anxiety.

In the 1990s, neurologist Antonio Damasio challenged the traditional split between thinking and feeling, arguing that emotions are rooted in bodily signals. His work showed that when this connection breaks, even logical decision-making falters. Meanwhile, neuroscientist Bud Craig mapped how the brain creates a live, inner dashboard of the body—tracking everything from energy levels to hidden threats—much like a starship captain monitors critical systems. This hidden sense, it turns out, shapes how we navigate the world.

Bengaluru-based Sarvam AI has secured $234 million in a Series B funding round, propelling its valuation to $1.5 billion and making it India’s newest AI unicorn. The round was led by HCLTech, which contributed $150 million, with additional participation from Bessemer Venture Partners and existing backers Khosla Ventures and Peak XV Partners. The startup aims to raise a total of $300 million in this round, marking a significant leap from the $41 million it raised across its seed and Series A rounds over two years ago.

The investment reflects a growing global push for sovereign AI capabilities, as nations and corporations seek greater control over critical AI technologies and computing infrastructure. Sarvam specializes in building full-stack AI solutions, from model development to enterprise applications, with a focus on Indian languages and use cases. Its models are deployed across banking, insurance, government services, and defense, and the partnership with HCLTech will combine Sarvam’s AI models with HCLTech’s enterprise relationships and engineering workforce to create tailored products for businesses and governments.

India has emerged as a key AI market, with OpenAI and Anthropic ranking it as their second-largest market after the U.S. However, high computing costs and limited capital have hindered Indian startups from competing with well-funded U.S. and Chinese rivals. Sarvam is among a handful of companies building homegrown foundation models. The urgency of AI sovereignty was underscored recently when Anthropic restricted access to its latest models following a U.S. government order, highlighting the risks of relying on foreign providers.

With the fresh funding, Sarvam plans to advance research into next-generation AI models focused on agentic, coding, and cybersecurity applications, while expanding computing infrastructure for large-scale deployments. The startup’s conversational AI platform already handles over 2 million daily interactions, its inference platform processes 10 million API calls daily, and its speech models transcribe 500,000 hours of audio monthly. Notable deployments include collecting data from 17 million farmers for India’s Ministry of Agriculture and supporting policy renewals for 45 million policyholders through a voice campaign for a leading insurer.

A cybersecurity startup called NewCore has emerged from stealth with $66 million in seed funding, aiming to solve a growing problem for companies deploying AI agents: how to manage their identities and access controls at scale. The round was led by Cyberstarts, with Index Ventures and Evolution Equity Partners participating, valuing the company at $300 million. As businesses increasingly treat AI agents as employees rather than tools, NewCore argues that existing identity platforms are not built to handle the complexity of a mixed workforce of humans and software.

NewCore’s platform treats AI agents as first-class identities, giving them their own permissions, life cycles, and revocation controls, separate from traditional service accounts. Co-founder and CEO Zohar Alon, who previously founded Dome9, says the idea took shape in 2023 when he saw a company paying a huge bill to an established identity provider but still unhappy with the service. He believes the market has grown stagnant, with vendors like Okta and Microsoft adding AI features as an afterthought rather than building for a future where digital workers operate alongside people.

The startup uses a “split-key” architecture to protect identity credentials and offers an “Agentic Skill” package for coding assistants like Claude Code and Codex, allowing them to access systems as managed identities. Employees can also use NewCore’s mobile app to grant or revoke access for AI agents on the go. With this funding, NewCore plans to scale its solution for enterprises preparing for a workforce that includes both human and AI participants.

For the first time, an Earth observation satellite has autonomously identified objects of interest without human assistance. In April, the YAM-9 spacecraft, built by Loft Orbital, used a vision-language model from Google DeepMind to analyze sensor data in orbit. The AI, running on a specialized Nvidia chip, responded to natural language queries by locating areas where human development meets natural terrain and identifying infrastructure near railway hubs. This marks a shift from traditional satellite operations, which require downloading massive datasets for ground-based analysis.

The demonstration relied on Gemma 3, a vision-language model designed for edge computing on limited hardware. NASA’s Jet Propulsion Laboratory developed the software package that integrated the model into the satellite. By processing data in space, the system reduces the volume of raw information sent to Earth, making satellite sensors more efficient. Loft Orbital’s head of AI, Paul Lasserre, described the capability as enabling always-on patrol layers that can monitor borders and alert users to suspicious activity.

This milestone has both immediate and long-term implications. In the near term, it allows satellites to triage data on orbit, cutting down the workload for human analysts. Longer term, it demonstrates the potential for larger-scale AI infrastructure in space. Loft Orbital operates 12 spacecraft and aims to build a constellation of 50 to 100 satellites for real-time global coverage. Other companies, including Planet Labs and Kepler Communications, are also exploring similar AI applications in orbit, signaling a broader industry shift toward autonomous space-based intelligence.

SpaceX’s record-breaking public debut this week, which catapulted Elon Musk to trillionaire status, has ignited a frenzy in the AI sector. While the company is known for space exploration, its costly AI ventures are now taking center stage. Rivals OpenAI and Anthropic have also filed confidentially to go public, setting the stage for a potentially historic summer of stock market launches, as discussed on the latest TechCrunch Equity podcast. The ripple effects are already reshaping market dynamics, with startups racing to capitalize on the momentum by pursuing orbital data centers and other AI-adjacent projects.

Analysts note that this wave signals a broader shift in public market priorities. The old FAANG grouping—Facebook, Amazon, Apple, Netflix, Google—is being replaced by what some call MANGOS: Meta, Anthropic, NVIDIA, Google, OpenAI, and SpaceX. This transition reflects a move away from consumer social networks toward AI labs and deep-tech innovators. The sheer scale of capital flowing into these sectors is unprecedented, and the upcoming IPOs will test whether public markets can absorb such concentrated, high-stakes investments.

For reporters, the summer promises a deluge of SEC filings and intense scrutiny. SpaceX’s offering not only absorbs vast sums but also challenges norms around corporate control, given Musk’s outsized influence. As other AI companies prepare to follow suit, the question remains: how much will they emulate this model? The answer could redefine the landscape for tech IPOs, with implications far beyond the headline-grabbing trillionaire milestone.