All Stories — Page 121
Every story published on RuntimeWire, newest first.
- Cyera signs $1 billion LOI to buy Oasis Security for agent identity controls
The proposed cash-and-stock acquisition would add Danny Brickman and Amit Zimerman's non-human identity controls to Cyera's data-security platform.
- Alibaba offers 2.1 billion QwenCloud tokens for agent feedback
The Qwen Growth Plan asks developers to submit successful and failed Qwen3.8-Max-Preview agent tasks by August 7th.
- Weave raises $13.5 million to measure the payoff from AI coding
Adam Cohen and Andrew Churchill are turning engineering measurement into a finance tool as companies scrutinize rising AI bills.
- Abby Care reaches eight states by turning family caregivers into paid employees
Founder Havi Nguyen's Medicaid know-how has helped Abby Care reach eight states, $15 million in 2025 revenue and a reported $225 million valuation.
- Hush Security raises $30M to govern AI agents at runtime
Akamai joined Battery Ventures and YL Ventures in backing the Meta Networks veterans, who plan to expand engineering and US sales.
- ZuriQ raises $25.5 million to scale its 2D trapped-ion quantum chip
Pavel Hrmo and his ETH Zurich co-founders must turn a nine-ion demonstrator into processors supporting hundreds of qubits.
- Fermion Research publishes 3.88 GB Neutrino-1 8B for local inference
The Qwen3-derived model packages 8.19 billion parameters in a 3.88 GB container using Fermion's packed ternary-family weight format.
- AGIBOT showcases A3 humanoid for stage performances and 100-robot fleets
The 121-pound biped pairs a 12 kW peak output with software for creating and synchronizing performances.
- Fermisense says a $500 Qwen fine-tune beat frontier models on catalog review
The 9B specialist scored 87.3% on Fermisense's simulation, though the test used 200 validation episodes and lacks independent replication.
- INFiLED launches dual-sided GeminiLED display for public spaces
The slim cabinet runs separate content on each face, giving airports, retailers and corporate venues one screen for two-way foot traffic.
- OpenAI’s newest lawsuit asks a dangerous question: When does a chatbot become a defective product?
A California man says GPT-4o reinforced his belief that he was Jesus, impersonated God and encouraged him as he attempted suicide. His 57-page complaint turns a horrifying chat transcript into a test of product liability, AI autonomy, medical licensing and whether long-context safety failures are engineering defects rather than unpredictable model behavior.
- Telnyx adds Moonshot's 2.8 trillion-parameter Kimi K3 to its inference API
David Casem is betting Telnyx can pair a frontier-scale open-weight model with the network, voice and messaging infrastructure agents need.
- Alleged OpenAI codenames Zinc and Magnesium surface in Design Arena
OpenAI's public GPT-5.6 materials identify only Sol, Terra and Luna, leaving the disabled entries unconfirmed.
- Baseten ships Rust tokenizer it says is 18x faster for Kimi K3
Michael Feil's new package targets a CPU bottleneck that grows as agent prompts approach Kimi K3's one-million-token limit.
- Depthfirst reports 105 flaws across Ruby projects with 8.6 billion downloads
The AI security startup says 62 findings are acknowledged by maintainers, after related flaws enabled code execution in self-managed GitLab.
- Simbryo discloses $5.25M round for organoid-based fertility diagnostics
Yael Katz's biotech sold the offering to 24 investors as it began marketing an assay designed to predict embryo implantation outcomes.
- Everything we know about how Anthropic's Project Panama scanned millions of books
Court records detail millions of pirated downloads, millions of purchased print books, a destructive scanning workflow and the 482,460-title settlement that followed.
- Dowe pitches one-source compiler for server, web and native apps
The project uses compiler-enforced contracts to contain AI code drift across server, web, desktop, Android, and iOS.
- Microsoft releases open browser agent post-trained on 64 B200s for six days
Fara1.5-27B uses screenshots instead of the DOM and ships under an MIT license, giving developers control over Microsoft's browser agent.
- Nativ v0.1.0 connects local Mac models to coding agents and image tools
Prince Canuma's open-source macOS app adds image generation, developer-tool integrations and live hardware monitoring.
- AMD releases Instella-MoE to showcase end-to-end training on Instinct GPUs
The 16B-parameter model activates 2.8B parameters per token and ships with six checkpoints, training recipes and code.
- BOSS Zhipin releases 3B agent model with a reused transformer stack
The Apache-2.0 model uses a two-pass Looped Transformer and was trained on 28 trillion tokens for coding, office and tool workflows.
- Anthropic rejects an open-weights ban while backing tests for frontier models
Dario Amodei is separating open access from the chip controls, anti-distillation rules and capability tests Anthropic wants Washington to adopt.
- OpenComputer launches managed deployment for long-running AI agents
OpenComputer's new Agent Deploy tool packages prompts into persistent, steerable agents with permanent URLs and managed infrastructure.