AI Daily Briefing — 2026-05-31

Groq, the AI chip startup known for its specialized hardware, is reportedly in the process of raising $650 million from its existing investor base, according to sources familiar with the matter. The fundraising effort comes just months after the company struck a unique $20 billion agreement with Nvidia that involved licensing its technology and the departure of several senior executives to the chip giant, a deal that was structured as a technology and talent transfer rather than a full acquisition.

The new capital is intended to fuel the expansion of Groq's inference cloud business, which allows developers and enterprises to run applications that require significant processing power after an AI prompt. This inference stage has become a critical bottleneck in the AI industry, with demand currently outstripping the need for model training. The company is currently being led by interim CEO Adam Winter and CFO Matt Eng as it pivots toward this growing market.

According to Axios, the funding round appears to be all but secured, with existing backers Disruptive and Infinitium agreeing to cover the full amount if other investors choose not to participate. This commitment underscores the confidence in Groq's strategic shift, even as the startup navigates the aftermath of its unprecedented arrangement with Nvidia.

GitHub Copilot is overhauling its billing model, moving from a flat subscription fee to a token-based system that could dramatically raise costs for many users. The change, effective June 1, charges developers based on the number of tokens consumed rather than a fixed monthly rate. While large enterprises may absorb the increase, smaller companies and independent workers are voicing alarm over potential budget strain.

On social platforms like Reddit and X, developers have shared stark examples of the financial impact. One user reported their monthly bill jumping from $29 to nearly $750, while another saw costs soar from $50 to around $3,000. Critics have labeled the new pricing “a joke” and “ridiculous,” with some announcing plans to cancel their subscriptions. However, not all feedback is negative—some argue that heavy token usage reflects poor coding habits rather than legitimate need.

Defenders of the new model suggest that developers who use Copilot efficiently will see minimal overage. They contend that those facing extreme bills are likely “vibe-coding” with excessive iterations. Others question the sustainability of the old pricing, speculating that Microsoft was subsidizing heavy users. A segment of the community also faults Microsoft for encouraging indiscriminate chatbot use before abruptly changing the rules.

TechCrunch AI reached out to Microsoft for comment but received no response by publication time. The debate highlights the tension between AI tool affordability and the costs of supporting advanced features.

Meta is reportedly working on an artificial intelligence-powered pendant, according to an internal memo reviewed by The Information. The device is expected to enter testing within the next year and builds on technology from Limitless, an AI startup Meta acquired in late 2025. Limitless had previously developed a pendant that could be clipped to clothing or worn as a necklace to record conversations. At the time of the acquisition, Meta stated the move would help accelerate its efforts in building AI-enabled wearables.

Earlier attempts at AI wearables have struggled to gain traction with consumers, often hampered by privacy concerns, poor marketing, or a lack of practical utility. However, companies like OpenAI continue to push forward in the space. The memo also indicates Meta plans to expand its line of AI glasses and introduce a business subscription service called Wearables for Work.

These initiatives are part of a broader strategy to turn around the fortunes of Meta's Reality Labs division, which posted a $4 billion loss in the first quarter of this year. TechCrunch AI has reached out to Meta for comment on the reported plans.

Google’s latest AI assistant, Gemini Spark, is designed to run around the clock on cloud-based virtual machines, handling tasks like summarizing emails, organizing spreadsheets, and managing to-do lists. Unveiled at Google’s developer conference in May, CEO Sundar Pichai pitched it as a more accessible alternative to always-on AI systems that require users to keep their laptops awake. The assistant integrates tightly with Google’s productivity suite, including Gmail, Calendar, Docs, and Sheets, positioning itself as a tool for streamlining work-adjacent chores rather than personal errands.

Despite Google’s struggle to present Spark as a must-have for personal use—suggesting features like daily email recaps or weekend planners based on calendar blocks—early testing reveals real utility. In one trial, the assistant helped identify weekly deals and clip coupons for a local drugstore trip, even recommending ways to stack promo codes for online orders. The results were practical, though not flawless: one suggested promo code turned out to be invalid.

Overall, Gemini Spark delivers a surprisingly useful consumer AI experience, even if it doesn’t quite justify its own distinct brand. Its strength lies in handling mundane digital tasks efficiently, but users may still encounter the occasional glitch that reminds them AI isn’t infallible.

As the browser wars intensify, a wave of innovative alternatives is emerging to challenge the dominance of Chrome and Safari. These new contenders are leveraging artificial intelligence, open-source principles, and a focus on user well-being to carve out their niche. From AI-powered assistants to privacy-focused tools, the landscape is shifting rapidly, offering users more choice than ever before.

Among the most notable newcomers are AI-centric browsers like Perplexity's Comet, which functions as a chatbot-based search engine capable of summarizing emails and sending calendar invites. The Browser Company's Dia offers a similar experience, using AI to help users navigate the web and perform tasks based on their browsing history. Opera's Neon and OpenAI's Atlas further expand this trend, with Atlas allowing users to interact with ChatGPT directly within search results.

For those prioritizing privacy and customization, open-source browsers remain a strong option. Meanwhile, a new category called "mindful browsers" is gaining traction, designed to enhance user well-being by reducing distractions and promoting healthier online habits. Startups like Aside, backed by Y Combinator, are also pushing boundaries with browser-native automation that can complete tasks autonomously.

As these alternatives gain momentum, the browser market is becoming more dynamic than ever. Whether users seek AI integration, enhanced privacy, or a more focused browsing experience, 2026 offers a wealth of options beyond the traditional giants.