Category: crypto
-

Google Chrome secretly installs 4GB AI model
Google Chrome has been installing a 4GB AI model called Gemini Nano on users’ devices without their consent, as discovered by researcher Alexander Hanff. The model reinstalls itself automatically if deleted, and Chrome does not offer an opt-out prompt during installation. This practice may violate EU privacy laws and raises legal and environmental concerns.
-

Kraken sues Etana over $25m client fund theft
Kraken’s parent company, Payward, filed a complaint alleging that Etana Custody and its CEO misappropriated over $25 million in client funds through a Ponzi-like scheme. Etana entered liquidation with significant liabilities, including a $16 million shortfall from risky bets. The case highlights issues in the crypto custody industry and is being closely watched for potential…
-

Kalshi valuation hits $22bn after $1bn Series F
Kalshi’s valuation has reached $22bn after a $1bn Series F round led by Coatue, doubling its worth in five months. The New York-based prediction market platform reported 800% growth in institutional trading volume and $1.5bn in annualized revenue. Despite regulatory challenges, Kalshi aims to expand its product suite and user base.
-

SIREN surges 22% but 4H chart flashes reversal
SIREN price on Binance perpetuals surged 22% on May 8, reaching $1.2965 before a 4H reversal indicated seller resistance. The daily MA ribbon is bullish, but the 4H MACD issued a caution flag. SIREN experienced sharp intraday reversals in the past, with supply concentration concerns remaining an issue.
-

Anthropic raise eyes $900bn valuation in summer
Anthropic is in talks to raise up to $50bn at a valuation of $900bn, potentially surpassing OpenAI as the most valuable private AI company. The company’s revenue has skyrocketed to over $45bn annually, largely driven by Claude Code. Investors, including Amazon, are showing interest in the raise, with a possible IPO in the future.
-

Meta’s Muse Spark ends its open-source AI era
Meta launched Muse Spark on April 8, its first fully closed AI model, moving away from its open-source Llama strategy. The model, developed by Alexandr Wang’s team after a $14.3bn investment in Scale AI, is proprietary with no open weights. It offers multimodal capabilities and is being marketed as a personal health reasoning tool. The…
-

JPMorgan makes AI core infrastructure spending
JPMorgan has reclassified its $2bn annual AI budget as core infrastructure, alongside data centers and cybersecurity. The bank has already generated $2bn in operational savings and runs over 500 active AI use cases, including fraud detection. JPMorgan is also investing in digital assets and using AI to predict liquidity needs.
-

OpenAI IPO targets late 2026 as revenue hits $25bn
OpenAI has reached over $25 billion in annual revenue and is preparing for an IPO in late 2026. The company has seen significant growth driven by ChatGPT subscriptions and enterprise adoption. With annual cash burn projected to reach $57 billion by 2027, going public is seen as a financial necessity.
-

US court rules AI ads make Meta liable for fraud
A US court found that Meta’s AI ads tools materially contributed to fraudulent investment content, removing Section 230 immunity and exposing the platform to securities fraud claims. This ruling could have implications for other platforms that use generative AI in advertising, potentially leading to increased securities fraud exposure.
-

Virginia redistricting vote struck down 4-3
Virginia redistricting referendum was struck down 4-3 by the state Supreme Court on May 8, with Democrats immediately filing to appeal to SCOTUS. The ruling found that the amendment violated procedural requirements and Democratic lawmakers had conducted the first vote after early voting had already begun, rendering the referendum null and void.