All Stories — Page 6
Every story published on RuntimeWire, newest first.
- VN.AI puts AI filmmaking and rights records in one production workflow
VN.AI's new platform tracks human edits, scans outputs for potential IP conflicts and brings work from outside tools into one record.
- Together AI adds DeepSeek's coding update to its API for agent workloads
DeepSeek released V4.1 Flash 1017 on October 17th; Together AI said it was live through its API on October 19th. The listing prices it at $0.30 per million input tokens and $1.20 per million output tokens.
- RuntimeWire marks five months with an AI newsroom run by one founder
RuntimeWire founder Ryan Merket says Cloudflare logged 4,705,949 unique IPs from August 18th to October 16th, a striking traffic figure for a five-month-old AI newsroom.
- Treg's open-source AEO toolkit finds sources cited by AI answers
Treg says the toolkit records sites cited by AI systems and identifies their authors; its broader product also gives agents access to third-party tools.
- CNTXT AI is opening its first U.S. office in Milpitas
Co-founder Mohammad Abu Sheikh is leading the UAE-based startup's expansion. CNTXT AI says it has $300 million in "committed revenues," mostly from U.S. buyers, and plans 15 to 25 commercial hires in its first year.
- SandboxAQ adds agent security and ROI tools beyond token counting
SandboxAQ, led by founder Jack Hidary, adds agent discovery, security testing, production guardrails and cost analysis, with a free plan for up to five managed agents.
- MiniMax says M3.1 Flash makes online inference nearly three times faster and cuts costs 70%
Founder Yan Junjie is pushing the model into coding and agent workflows; MiniMax's speed, price and benchmark figures are company-reported.
- Accenture and Google launch Gemini agent-building program for federal agencies
Agencies can start with hands-on workshops at Accenture's Washington, D.C., Forge and Prototyping Lab, then work with embedded engineers on Gemini agents. The announcement names no pilot agency and specifies no launchpad-level security authorization, pricing, or procurement route.
- Higgsfield's CLI runs AI generations from scripts on its existing credit meter
Higgsfield says its open-source CLI can call more than 35 models, automate batch jobs and charge generations to existing plan credits.
- SAP agrees to acquire TechWolf, adding a test case for Ghent's startup flywheel
TechWolf CEO Andreas De Neve reported US ARR rising from $1 million to $15 million in 18 months and models serving millions of people. Co-founder Jeroen Van Hautte says the founder cycle matters to Ghent; local examples show a broader pattern, not a proven TechWolf effect.
- STROT compiles LLM instructions into locally run decision workflows
Amit Rath's open-source project puts a large model in the setup step and a small local model on the job, while its Pac-Man demo shows the limits of that split.
- CyberSN counts 26% fewer US cyber postings as AI security roles climb
Its January-September tally hit a four-year low, even as AI Security Engineer postings rose to 2,409; the figures count listings, not hires.
- Together AI researchers release agents that shift reasoning out of text
The team reports lower token use and faster inference across 11 benchmarks, while its cost savings remain estimates that exclude prompt processing and tool calls.
- Data World builds auditable synthetic worlds for data agents
A paper led by Zhanli Li and researchers at Meituan and universities uses explicit causal rules to generate messy files and score data agents against the hidden record.
- Air Force tests Anduril's Fury under F-22 pilot command
The Air Force said the test took place earlier in October and announced it on October 16th; available reporting does not give the test's exact date. Fury followed tactical instructions from the F-22 pilot during simulated combat.
- EG Secure patches exploited SiteGuard flaw that can grant full server control
EG Secure's default-enabled statistics feature leaves specified Apache httpd, Nginx and IIS releases vulnerable; administrators can upgrade, apply its patch or disable statistics after saving diagnostics.
- Anthropic proposes putting 5% of its shares in a public-benefit fund
The proposal needs stockholder approval. Anthropic says the fund could receive more equity as Anthropic grows and estimates roughly 20% of its shares will support public benefit.
- A Plex honeypot was hijacked after an unauthenticated exploit chain surfaced
The operator reported a cryptominer and proxy on one old server; Plex urged upgrades in September, and the public chain targets versions before 1.43.3.10861.
- Vals AI's refund test finds Grok sent 187 messages in 15 days
The study, published on October 18th, put six personal-assistant agents in a simulated refund dispute and tracked what they did after the claim deadline passed.
- Disney estimates $180M in pay and retention for Character.AI hires
The filing estimates about $180M in compensation and retention for roughly 65 employees. Disney's separate $20M stalking-horse license bid remains subject to court approval and higher offers.
- Startup Spotlight: Delineate turns drug papers and plots into datasets for pharma models
Founded by Emily Nieves and Jawad Iqbal, Delineate says it has built datasets for pharma clients from hundreds of trials and thousands of published plots. Public customer counts differ, and its performance claims remain company-reported.
- Startup Spotlight: Hamming AI turns voice-agent calls into repeatable QA tests
Hamming AI founder Sumanyu Sharma, a former Tesla and Citizen data leader, built a platform that simulates calls, tests audio and telephony edge cases, and monitors agents in production. Hamming raised $3.8 million in 2024; its usage figures are company-reported, while revenue and retention remain undisclosed.
- Startup Spotlight: LunaBill automates claims follow-up after Suhail Parry worked the billing queue
After spending a year inside a medical billing operation, Suhail Parry built LunaBill's agents to work claims across payer portals, calls, denials and appeals.
- Anthropic's draft AI safety bill lets only Commerce seek a release halt
The draft sets a threshold of more than 50 people killed or seriously injured, or more than $1 billion in damages, before Commerce can ask a court to halt a release.