Tech·4,610 stories
Tech News — Page 204
Vertex AI Default Permissions Expose Cloud Data, Artifacts
Palo Alto Networks Unit 42 disclosed a security flaw in Google Cloud's Vertex AI platform in which default service agent permissions let deployed AI agents expose cloud storage credentials and access
Anthropic Claude Code Leak Shows 512k Lines & Pet
A packaging error in Anthropic’s Claude Code update exposed over half a million lines of its source code, revealing unreleased features like a Tamagotchi‑style pet and an always‑on agent, prompting a
OpenAI Raises $122B, Adds $3B Retail Investment
OpenAI secured a $122 billion funding round at an $852 billion valuation, bolstered by $3 billion from retail investors and a $4.7 billion credit line, as it ramps up revenue, user growth, and ad‑base
iOS 26.4 Enables Voice‑Only ChatGPT on CarPlay
Apple’s iOS 26.4 update now lets users launch the ChatGPT app on CarPlay, but the interface is voice‑only, requires tapping to start a conversation, and complies with Apple’s developer rule against on
iOS 26.5 beta 1 adds encrypted RCS messaging
Apple has released iOS 26.5 beta 1, introducing significant updates including end-to-end encrypted RCS messaging in the Messages app, as reported by 9to5Mac, and new features for Apple Watch with watc
OpenAI Closes $122B Round at $852B Valuation
OpenAI closed a record $122 billion funding round at an $852 billion post-money valuation, led by SoftBank and Andreessen Horowitz, expanding on a previously announced $110B target and drawing $3B fro
Axios npm package compromised in supply chain attack
A supply chain attack compromised the Axios npm package — one of JavaScript's most widely used HTTP clients with roughly 100 million weekly downloads — by injecting a malicious dependency that deploys
Claude Code users hit usage limits faster
Anthropic's Claude Code users are encountering usage limits much faster than anticipated, indicating high demand for its coding capabilities. This rapid consumption suggests developers are heavily lev
Microsoft, Amazon, OpenAI launch AI health tools as
The rapid proliferation of AI health tools from **Microsoft, Amazon, and OpenAI** is meeting a clear demand for accessible medical advice, yet concerns are mounting over their lack of external evaluat
AI benchmarks fail in real-world teams, researcher argues
A UCL professor argues that current AI benchmarks mislead organizations about real-world performance and proposes evaluating AI systems within human teams and workflows over longer time horizons.
Apple at 50: Integration Moat Meets AI Aggregation
As Apple turns 50, an analysis argues its half-century of hardware-software integration uniquely positions it for the AI era — even as the company trails peers in AI server spending and has yet to shi
Inc42 Launches Programs D2CX, ManagementX, Foundations
Inc42 rolls out three D2C-focused programs—D2CX, ManagementX, and D2CX Foundations—to equip Indian entrepreneurs with expert guidance, growth tactics, and leadership training, promising to accelerate
OpenAI Kills Sora, Burns $1M/Day, Ditches Disney Deal
OpenAI is shutting down Sora, its AI video generator launched just six months ago, redirecting compute to other tasks after the product reportedly burned $1 million a day, lost half its users, and for
LiteLLM Drops Delve After Malware, Will Re-Certify With Vanta
LiteLLM is dropping compliance startup Delve and re-certifying with Vanta after credential-stealing malware hit its open source AI gateway and a whistleblower accused Delve of fabricating compliance d
Ex-Coatue Investor's Sycamore Raises $65M Seed
Sycamore, a startup building an enterprise AI agent orchestration platform, raised a $65 million seed round led by Coatue and Lightspeed, with a high-profile angel list including former OpenAI chief s
76% of Americans distrust AI as daily use grows
A Quinnipiac University poll reveals a striking paradox in American attitudes toward AI: usage is climbing but trust is cratering, with 76% of respondents saying they rarely or only sometimes trust th
OpenAI patches ChatGPT data exfiltration flaw
Check Point discovered a ChatGPT vulnerability that silently exfiltrated user messages, uploaded files, and sensitive data through a hidden DNS channel in the AI's Linux runtime, bypassing all built-i
Bluesky's Attie Blocked 125,000 Times in Days
Bluesky's new AI assistant Attie, which lets users build custom feeds within its AT Protocol, was blocked by roughly 125,000 users within days of its weekend launch — making it the second most blocked
OpenAI Abandons Erotica Plans Amid Safety Fears
OpenAI abandoned plans to launch AI-generated erotica for verified adults last week, citing safety concerns and investor alarm after rival xAI's Grok produced illegal child sexual abuse material.
AI Health Chatbots Launch Without Independent Testing
Microsoft, Amazon, and OpenAI have all rolled out consumer-facing AI health chatbots in recent months, but researchers warn the tools ship to the public without independent safety testing.
Get the Tech newsletter
Curated tech stories, every morning. Free.
No spam. Unsubscribe anytime.