Only Tue, October 20, 2026 · All dates →
Ornith-1.5-122B released: self-improving model built on QwenOrnith AI released Ornith-1.5-122B, which extends self-improvement from Ornith-1.0 (built on Qwen3.5 and Gemma4) to jointly optimize task generation, scaffold construction, and solution rollouts via reinforcement learning. The model is available in GGUF format on Hugging Face.
Spotify pulled 50M AI slop and spam tracks from recommendationsSpotify said it removed more than 50 million tracks deemed AI slop or spam from recommendations over the past year and adjusted its system to favor human-made music. It reports fewer listener complaints but has not mandated labels on all AI tracks.
Figma co-founder says MCP is not a perfect protocolAt its first Config conference in India, Figma announced 20 new MCP partners, including Notion, Linear and Mistral. Co-founder Dylan Field said MCP is imperfect and favors frontier labs, and that Figma won't expose all its data via MCP due to abuse risk.
Teen AI prodigy Nathan Chen said to join US startup Discovery LoopTeen AI prodigy Nathan Chen is reportedly joining US startup Discovery Loop, according to South China Morning Post. His role was not disclosed.
Adobe Lightroom mobile adds one-tap AI Open Eyes featureAdobe said Lightroom on iOS, iPadOS and Android now offers a generative AI Open Eyes Quick Action: when the app detects closed eyes, an icon appears over the photo and a single tap opens them. Similar capability has existed in Photoshop Elements for nearly a decade.
Citrini Bets on Life Sciences After Testing Claude AI Drug StudyCitrini Research, after testing Anthropic's Claude Science, argues that AI-driven drug discovery will accelerate, benefiting life sciences companies. The firm is bullish on the sector's investment opportunities.
Report: legal and compliance AI market to reach $151B by 2030The Business Research Company projects the legal and compliance AI market to grow from $29.68B in 2025 to $41.19B in 2026, a 38.8% CAGR, and reach $151B by 2030. Growing regulatory complexity and document volumes are cited as key drivers.
Eka Robotics raises $150M Series A at $1.45B valuationEka Robotics, founded by MIT professor Pulkit Agrawal and former DeepMind researcher Tuomas Haarnoja, raised a $150M Series A led by a16z and F-Prime at a $1.45B post-money valuation. Its force-sensitive gripper and deep-learning model pick up varied objects without retraining, at about 1,500 items per hour with over 95% success.
Anthropic and Musk Unveil Big Philanthropic Plans, Beneficiaries UnclearAnthropic and Elon Musk announced major philanthropic plans, but the report questions who actually benefits and how the money will be used. It focuses on transparency and allocation of AI firms' philanthropic commitments.
RBC Weighs Blocking Meta's AI Agent Muse, CEO SaysRoyal Bank of Canada CEO Dave McKay said the bank is considering blocking Meta's AI agent Muse from accessing customer information to protect consumers and limit liability. The move highlights financial institutions' concerns over third-party AI data access.
CVPR 2027 author kit not yet posted onlineA Reddit user asked whether anyone has the CVPR 2027 author kit, noting the official site doesn't appear to have posted it yet, unlike in previous years.
Meta Muse share links can be enumerated, leaking user dataSecurity firm Claroc found that share links created by Meta's Muse AI agent can be enumerated; researchers listed 6,800 links and found clearly non-public personal and business information in some. CEO Mark Zuckerberg had said last month that Muse was “built from the ground up for privacy and security.”
Army Cyber Command Develops Self-Spreading AI CyberweaponsAccording to internal documents, the U.S. Army Cyber Command is developing AI agents codenamed ECHOTRIBBLE that can move through target networks like a worm, 'neutralize' targets, and even reach air-gapped systems. Experts warn of potential collateral damage.
China sets up state-owned data group to power AI ambitionsChina has established a state-owned data group directly administered by the central government to strengthen data infrastructure and bolster technological development for its AI ambitions. The move shows Beijing treating data resources as core to AI competition.
Salesforce to acquire marketing platform Braze for $5.7BSalesforce has agreed to acquire Braze, an AI-powered marketing platform, for almost $5.7 billion in enterprise value, expanding its marketing cloud capabilities. It marks another major deal by Salesforce in AI-driven marketing.
Alphabet as an AI investment alternative benefiting from OpenAI and AnthropicA Motley Fool piece argues Alphabet is a compelling AI investment before OpenAI and Anthropic go public. Google Cloud supplies compute to both rivals: OpenAI agreed in May 2025 to use Google Cloud capacity, while Anthropic uses Google's custom silicon more deeply.
AI adoption varies by industry, with finance second to techSurvey data from TripleTen shows uneven AI adoption: 48% of finance workers have been asked to use AI chatbots, versus 18% in healthcare and 25% in education. The slower pace outside tech reflects different workflow integration speeds rather than lagging behind, the article argues.
Who is Yuki Chen, the Chinese hacker who joined AI firm Z.ai?The South China Morning Post profiles Yuki Chen, a Chinese hacker who recently joined AI company Z.ai. The report is framed around who he is, but no body text was accessible.
Interpol warns of AI-powered deepfake fraud schemesInterpol warns of a rise in AI-enabled fraud that fakes faces, voices, and behavior to impersonate executives, friends, or relatives and extort money. It advises slowing down, calling back on a known number, using code words, enabling two-factor authentication, reporting fraud, and training staff.
AI token billing brings hidden costs, with one user at $28KToken-metered AI tools are becoming a new business expense. Business Insider reported in August 2026 that a Microsoft employee ran up $28,000 in AI charges in 28 days. The author urges users to watch后台 metering and turn on real-time notifications to avoid surprise bills.
Crunchbase: four AI firms took 65% of Q1 VC fundingCrunchbase data shows four companies raised $188B in Q1 2026, 65% of global venture investment, while only 20% went to non-AI firms and deal count fell 30%. The author, who built a multibillion-dollar company without VC, urges founders to prioritize customers and revenue over fundraising.
Fintech funding falls 21.4% in Q3 as deal count hits lowGlobal fintech startups raised $10.6B in Q3 2026, down 28.4% quarter over quarter and 21.4% year over year, Crunchbase data shows. Just 687 rounds were recorded, the lowest since early 2018 and down 31% YoY. Investors describe a bifurcated market favoring AI and leaders, with $100M+ rounds down 43.3%.
Abiologics Appoints Per Jr. Greisen as Chief AI OfficerAbiologics has appointed AI protein design pioneer Per Jr. Greisen as its Chief Artificial Intelligence Officer. He will lead the company's AI and machine learning strategy to advance its Synteins platform for synthetic D-protein therapeutics.
Yelp launches AI voice agent to check restaurant wait timesYelp added several AI features to its app, including 'Check for Me,' which lets a voice agent call restaurants on users' behalf to check wait times and text results back. Its assistant also gains OpenTable booking, preference memory, reminders, and custom lists.
Otter.ai raises $320M from General Catalyst for meeting toolMeeting-notes tool Otter.ai said it raised $320M from General Catalyst's Customer Value Fund, which supplies capital without equity dilution and earns returns based on marketing results. Otter.ai will use the funds for enterprise growth, go-to-market strategy, international expansion and product innovation.
Vals AI: DeepSeek V4.1 Flash beats GPT 6 Sol on cybersecurity taskVals AI found that DeepSeek V4.1 Flash outperformed GPT 6 Sol on a multi-step offensive cybersecurity task. Sol was tested via the Daybreak trusted access program.
Tecnalia Launches VAIBreaker Tool to Assess AI SecuritySpanish technology center Tecnalia has released VAIBreaker, a tool for assessing the security of artificial intelligence systems. It aims to help identify and mitigate AI security risks.
Amazon, Meta AI wearables raise privacy concernsBloomberg reports that AI wearables from Amazon and Meta are raising privacy concerns over always-listening capabilities. The always-on devices continue to draw scrutiny over how they collect and use user data.
ISTA-DASLab Uses Prefiller Model to Boost Qwen3.8-27B Low-Quant AccuracyISTA-DASLab demonstrated a prefiller model approach that recovers low-quant accuracy for Qwen3.8-27B while retaining low VRAM usage and fast decoding. It requires an Nvidia Blackwell GPU and shows no speed penalty for contexts longer than ~4k tokens.
Best practices for annotating 500 images to fine-tune Meta DETR?A developer seeks advice on efficiently annotating ~500 images for fine-tuning Meta DETR, considering tools like CVAT, Roboflow, Grounding DINO, and OWLv2 for AI-assisted annotation. The discussion focuses on balancing speed and accuracy, and workflows to ensure consistent bounding boxes and class labels.
The A.I. Daily