@openai : OpenAI says ChatGPT Voice can now be powered by GPT-6 Astra, Sol, and Luna, use plugins like email and calendar, and be used in ChatGPT Work on web and mobile — We heard you loud and clear. ChatGPT Voice can now: - Use plugins like your email, calendar, and Slack.…
Read sourceAnthropic said its Claude system identified a previously uncharacterised enzyme system in bacteriophage DNA after a long search through sequence data, which the company presented as its first result from a new wet lab.
Read briefAnthropic said its Claude system found a previously uncharacterised enzyme system in bacteriophages after searching genomic data, while researchers said its biological function is still unknown.
Read briefOpenAI updated ChatGPT Voice to use new GPT-6 Astra, Sol and Luna models and to reach email, calendar and Slack, while also extending voice use to ChatGPT Work on web and mobile.
Read briefGoogle launched Gemini 3.8 Flash TTS and Flash-Lite TTS, saying the models could create voices from text, clone voices from short samples and handle scripted dialogue through its developer tools.
Read briefYouTube said it was adding AI-assisted scripting, thumbnail selection, editing, translation and moderation tools for creators across Studio, Shorts and YouTube Create.
Read briefKate Clark / Wall Street Journal : Pilgrim, whose device combines air sampling and genomic sequencing to detect biological threats, raised a $25M seed led by Buckley at a $150M valuation — Pilgrim has raised $25 million and has developed a 50-pound device that detects…
Read sourceMarking two years of OpenAI Academy and bringing AI skills to even more communities.
Read sourceGoogle announced two Gemini text-to-speech models, said they supported voice creation and replication across more than 100 languages, and outlined availability through Google AI Studio, the Gemini API and Gemini Enterprise.
Read briefGoogle introduced Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS, adding custom voice design, voice replication and line-by-line direction, and said they were rolling out across its products and developer tools.
Read briefAnthropic employee Jackson Kernion said newer Claude models had been tuned more for maths, code and model-to-model explanation, which he said had left their prose harder for people to read.
Read briefAmazon said it was linking seller tools to Anthropic's Claude and its own Quick assistant, while adding background automations, a revised Seller Assistant and a free Quick Plus subscription for merchants.
Read briefMeta said its AI agent Muse drew more than 500,000 users in its first week in the US, while the company and others pointed to strong similarities with OpenClaw.
Read briefYouTube introduced a Custom Feeds button that uses large language models to build and save personalised homepage video feeds, with a wider rollout planned for US users.
Read briefThe London company said new funding would support EDEN, its biological AI models, and therapy programmes built from its own data collected with partners across more than 30 countries.
Read briefBasecamp Research raised $140m in a Series C led by S32, with Anthropic and other investors joining, and said it would use the money to train new AI models and advance therapeutics.
Read briefOpenAI is extending access to its Daybreak program to the Government of Ukraine to support the cyber defense of civilian infrastructure.
Read sourceOpenAI published a customer story on Ringg that said its GPT-5.6-based agents handled up to 65% of conversations themselves and cut spending by roughly 90% versus GPT-4.1.
Read briefAnthropic and OpenAI said their latest models showed lower rates of some unsafe behaviour in tests, while OpenAI also outlined plans for outside safety assessments.
Read briefOpenAI said it had hired Patreon co-founder Sam Yam to lead a new creator product team, with two former Patreon executives joining him and a first preview possible next week.
Read briefAirbnb said a new agreement gave its engineering and product development teams broader access to OpenAI frontier models, including GPT-6 Astra, through APIs and Amazon Bedrock on AWS.
Read briefGrab and OpenAI said they had launched GO Forward with AI, a regional training programme for 30,000 partner workers across Southeast Asia that will begin in Singapore and later expand to three other countries.
Read briefOpenAI introduced GPT-6 Sol and Luna at lower token prices, said the models matched earlier versions on cost-adjusted benchmarks, and an independent analysis found mixed score changes.
Read briefCisco Talos said a Windows malware sample named ClosedQuorum used four AI models to decide post-compromise actions automatically, with tie-breaks set by priority and some features missing in the analysed build.
Read briefMicrosoft said it disrupted EvilTokens, a device-code phishing service linked to more than 12,000 inbox compromises, with court authorisation and help from law enforcement and private-sector partners.
Read briefResearchers said a flaw in Bifrost’s management API let an unauthenticated attacker register an MCP client and run commands on the gateway server, with fixes available in later releases.
Read briefCheck Point said attackers had been exploiting two pre-authentication vulnerabilities in its Security Gateway and management web service, with one used as a zero-day since 23 July and fixes now advised.
Read briefPatchstack said attackers shifted from reconnaissance to file-writing attempts against CVE-2026-87902 after WordPress 7.1.2 shipped, while WordPress described the bug as a path-traversal issue that could lead to remote code execution.
Read briefAikido said four packages in HashiCorp’s registry carried Go malware that overlapped with Graphalgo, while other disclosures in npm and Rust-related activity described similar lures and command channels.
Read briefGambit said an ongoing campaign using three open-source AI tools had targeted online retailers since July, stolen more than 600,000 card details and planted skimmer malware on at least 119 sites.
Read briefCERT Polska said two RouterOS SSH flaws combined to let unauthenticated attackers reach administrative commands on exposed MikroTik routers, and logs showed possible exploitation before patches were released in September.
Read briefCisco Talos said the Windows malware CLOSEDQUORUM queried up to four commercial AI services to decide what to do next, though the sample it analysed did not fully function.
Read briefResearchers said attackers altered MemTensor packages on npm and PyPI to spread a Go implant called sckit, which stole credentials and secrets from developer systems and cloud accounts.
Read briefArista Networks said it fixed a zero-day in VeloCloud Orchestrator that attackers were already using, while CISA added the issue to its exploitation list and set a federal deadline.
Read briefDepthFirst said it found a Linux kernel use-after-free that could let a container reach host root, and it published exploit code while Ubuntu had not shipped fixes for affected releases.
Read briefF5 said a BIG-IP Access Policy Manager flaw was being exploited when the module acted as an OAuth authorisation server, and it issued hotfixes after CISA listed it in KEV.
Read briefVolexity said UTA0565 used fake websites and phishing emails to chain Chrome and Windows flaws in early September 2026, then delivered CLEANGULP through a BlueMoon exploit kit.
Read briefKaren Serobovich Vardanyan received a 24-month prison sentence and three years’ supervised release after pleading guilty to hacking U.S. organisations in Ryuk ransomware attacks and being extradited from Ukraine.
Read briefF5 said it released security updates for a BIG-IP APM vulnerability that was already being used in remote code execution attacks, and CISA set a Friday deadline for federal agencies.
Read briefVercel said a vulnerability in Next.js ImageResponse affected versions 16.2.0 through 16.3.5 on Node.js, could lead to server code execution through attacker-controlled SVG input, and was fixed in 16.3.6.
Read briefGreyNoise said a campaign it tracked from June 2026 used WordPress Core and ZyXEL flaws to breach organisations in 29 countries, steal records and collect device credentials from switches.
Read briefShinyHunters said it used an unpatched Oracle PeopleSoft flaw to reach FBI systems, take employee and applicant records, and deface the FBI Jobs site, while the account remained unverified.
Read briefCheck Point said attackers used a management server flaw in July and reported later attempts against a separate VPN issue that it had already patched in September.
Read briefReversingLabs said the package tw-pkgprobe-7731 mimicked a Twilio bug-bounty tool, while later versions probed Twilio-related systems and one release sought ACCOUNT_SID and AUTH_TOKEN from the environment.
Read briefA GitHub proof-of-concept called BigDiskBuster prevented Microsoft Defender from installing updates by consuming free disk space, while its author said no patch or independent confirmation was available.
Read briefArista said attackers were exploiting a CVSS 10.0 flaw in on-premises VeloCloud Orchestrator deployments that use certificate-based Edge authentication, and it had released fixes for some supported versions.
Read briefA flaw in Linux’s ARM64 KVM nested-virtualisation code let a guest reach freed host memory, and researcher Hyunwoo Kim said it could be used to escape to the host.
Read briefA researcher published technical details of CVE-2026-65660, a SharePoint Server flaw that Microsoft had described as spoofing, while the separate vulnerability record and the National Vulnerability Database treated it as code execution.
Read briefCheckmarx said the removed npm package indexed-btree moved its loader into normal runtime code, while the wider article also described a related Packagist campaign and defensive advice.
Read brief