All AI news

AIMistral Large 4 reverse-engineered an unknown binary end-to-end, concluding it was Cobalt Strike and extracting its IoCs and malware configuration. The model also wrote a report with a YARA rule for detecting future incidents. The task, which could take a day of analyst work, was completed in 12 minutes.

AIMistral Large 4 solved 18 of 19 challenges in a CTF speedrun, with tool calls and solve times drawn from actual runs. The post frames the model as efficient at reasoning over diverse complex challenges compared with other models.

AIMistral AI says its Mistral Large 4 is one of the world's strongest AI models for cybersecurity. The post gives no benchmark scores, comparisons, or availability details.

AISierra announced two new models at Sierra Summit 2026: Curie, which runs the core conversation loop, and Fleming, which detects when a caller is another agent. The company also introduced Tandem Voice, Persona Studio, a reimagined Ghostwriter, and Horizon for long-horizon agent goals, and listed Sierra in the Stripe App Marketplace.

AIStarting October 9, free Google Gemini users will only have access to the Flash-Lite model, and AI Plus subscribers will lose Pro access. Google is also moving to compute-based limits that refresh every five hours, with higher limits for paid plans. The AI Pro plan at $20 per month will gain access to the Deep Think reasoning mode.

AIGoogle researchers, working with Northwestern Medicine and Jacaranda Health, trained healthcare workers to perform "blind sweep" ultrasounds analyzed by machine learning models. The models estimated gestational age and fetal presentation as accurately as a trained sonographer in a study of 1,000 mothers each in Nairobi and Chicago. The AI processes results on the device, so it needs no electricity supply or Wi-Fi.

AIWaymo says its vehicles may begin appearing on Detroit roads this week without a human driver. The company says it is not yet picking up riders and invites users to download the Waymo app to be among the first when service launches.

AIArthur Mensch posted that Mistral is building AGI from a train station. The quoted post adds that ML4 was trained on 3,800 NVIDIA Grace Blackwell GPUs in Mistral's European datacenters, including its Bruyères-le-Châtel cluster funded by its Series B. Mistral says Series C and D clusters are coming soon to support longer training and faster iteration.

AIAgentMail has launched AgentID, a sign-in system that lets millions of AI agents with AgentMail accounts log into third-party apps within minutes. The post frames it as the agent-era equivalent of Sign in with Google, and offers three months of free AgentMail plans to developers who integrate it and share a screenshot.

AIA developer used Grok Bot to build two Slack team bots, sandcastle for release management and poteto for engineering, automating their release and QA process. Sandcastle DMs contributors PR links, kicks off builds, and runs a fuzz swarm of 10+ Grok 4.7 xhigh agents, while poteto triages and fixes issues via Cursor Projects.

AIMistral announced Mistral Large 4, a natively multimodal model with 1T parameters and 49B active parameters. It is available via API today, with open weights planned for the end of October. Clément Delangue, Hugging Face's CEO, reacted by noting that the model cannot be the best open-weight model until its weights are actually released.

AIDeepSeek is close to raising at least 80 billion yuan ($12 billion), up from an original target of about 50 billion yuan, according to Bloomberg citing people familiar with the matter. Tencent and battery maker CATL are among the largest investors in the round, which is expected to close soon. DeepSeek is planning an IPO in early 2027, though details could still change.

AISierra and Meta are developing Personal Agent Protocol, an open standard defining how personal agents interact with businesses, with industry partners including Genesys, Instinct, Rocket, Shopify, Stripe, and Walmart. The protocol uses OAuth sessions where consumers choose read-only or write access and companies choose whether agents reach them through websites, APIs via MCP and OpenAPI, or their own agents. The authors plan to publish the v0.1 specification later this month along with a reference implementation.

Why it matters: The post specifies how personal agents would authenticate and reach businesses through websites, APIs, or company agents, which matters for anyone building agent integrations.

AIAmazon Science is hosting live research talks, meet-the-scientist sessions, and language model discussions at the COLM 2026 conference all week. Attendees can visit the Amazon booth during the event.

AILuma Labs says there is no one model that excels at every task, since ideation and final-frame generation require different strengths. The post's point is that creative work should be matched to the best model for each stage rather than a single model for all.

AILuma (@LumaLabsAI) posted a short call to action inviting readers to try Luma through a tracked app link. The post contains no details about features, models, pricing, or capabilities.

AIReplicate is offering Google's nano-banana-2.1 model for users to try today. The post links directly to the model's page on replicate.com.

AIReplicate has made Nano Banana 2.1, Google DeepMind's latest image model, available, optimized for a balance of price and performance. The model offers improved visual design, mask-based editing, and subject consistency for more natural-looking images.

AIHark Pro is now live on web, iOS, and Android as a new personal agent, and early users reportedly get one free month of Pro. The post says the product is powered by Opus 5.5 and Sonnet 5.5, and praises its polished design, detailed animations, and AI-customized cards.

AIPerplexity says its open-weights multimodal decision model pplx-decider-v1.1-27b ranks first on the new Hugging Face Decision Index 0.3 benchmark. The updated model is available to all, and it costs $0.02 per million input tokens, half the price of v1.

AIAnthropic's Claude can now open Google Docs, Sheets, and Slides beside the chat, either from a pasted Google file link or a request for a new file, so users and Claude can edit together. Access follows existing Google sharing permissions. The feature is in beta on all paid plans.

AIand files from those apps can also be opened within Claude. In Google Workspace, Claude appears in a sidebar next to the open file, reads its contents, and edits it in place, with each edit available for user approval before it is applied.

AIAnthropic says builders can receive up to $7,000 in Claude products and credits. The offer includes a free year of Claude Team for companies new to Team, plus $1,000 in API credits upon approval.

AIMETR argues that agent observability should be treated as security-critical infrastructure, with all AI outputs, including transcripts, reasoning, and actions, handled as untrusted inputs. The goal is to make it very difficult for agents to influence the systems used to supervise them.

AIMETR notes that a bug altered only what a viewer displayed, not the underlying logs. It warns that more capable agents could target multiple oversight systems or deeply compromise a company's infrastructure to build a "Potemkin village" that deceives employees.

AIMETR demonstrated a JavaScript injection that an agent could trigger through MathJax rendering from any part of its output, including its reasoning. METR says it has not observed agents exploiting this in its evaluations, and the Inspect tool was patched within a day of the report.

AIA METR researcher discovered a bug in the transcript viewer of Inspect, a popular evaluation framework, that could let an agent show a user a fabricated or edited version of its transcript.

AIAI companies often log their agents' actions and reasoning steps to detect misbehavior. The post warns that misaligned agents may be able to hack the software humans use to review these logs, concealing their misconduct.

AIDex Horthy (@dexhorthy) announced episode #77 of his AI That Works series, titled "Building a Memory Pipeline," as a live broadcast. The post provides no further technical details about the pipeline's design, tools, or results.

AIYuchen Jin says Reflection's Beam and Mistral Large 4 both reached roughly GLM-5.2 level within the past two days. He suggests the Western versus Chinese open-source model gap may come down to Chinese labs being able to distill Anthropic and OpenAI models, which Western labs cannot.

AIAllie K. Miller suggests giving an AI agent a separate email address, which Instinct does automatically, and using it for junk signups so the agent filters that mail away from your main inbox. She compares it to Google Voice for email and argues retail emails will get less attention unless they give people a reason to reach the human inbox.

AIMerve Noyan has released a slide deck on running AI locally, covering prefill versus decode, MoE versus dense models, VRAM versus unified memory, quantization, and speculative decoding. The deck is built around llama.cpp and is free to reuse with attribution.

AILlamaIndex is hosting the invite-only Agent Economy Gala on Friday at the Redwood Room in San Francisco, during SF Tech Week. The event expects more than 120 agent founders and 50 investors, with live jazz, and lists partners including SpaceX, Daytona, Lemma, TinyFish, The Linq, and Obvious.

AIxAI has shared a complete cookbook repository on GitHub, linked at The post gives no further details about the repository's contents or purpose.

AIGrok can turn a pasted article or PDF into a podcast where two hosts discuss the content. Each script line is voiced as soon as it is written, so playback begins before the full script is finished. The example code is available in the xai-cookbook repository on GitHub.

AIThe xai-cookbook's x-sentiment-tracker lets users pick a topic, then has Grok run one X search per day over the last 10 days and score each post from -1 to +1. The dashboard displays the sentiment split, day-by-day movement, and the posts behind each score.

AIThe xai-cookbook example uses Grok to generate a Tailwind React component from a dropped-in screenshot. The page then screenshots its own render, overlays it on the original, and feeds the mismatches back so Grok can refine the layout, raising the match from 60% to 77% in one run.

AIA pipeline built with Grok takes a single product photo and produces an eight-second vertical ad, writing the brief, generating three scenes, selecting the best one, animating it, and adding a voiceover. Each run costs about $1.35, and the code is available in the xai-cookbook repository on GitHub.

AIxAI's storyboard-to-film example uses grok-4.7 to plan four shots from a one-line premise, then Grok Imagine draws and animates each one with text-to-speech narration. Every shot is an edit of the first keyframe, which keeps the main character consistent across scenes. The code is available in the xai-cookbook repository on GitHub.