标签：#tech

New EU rules stipulate that people must be told when they’re interacting with AI or looking at AI-generated or -edited content, leading to fear of “disclosure fatigue.”

The Open-Weight Inflection Point: Kimi K3, Claude Opus 5, and Microsoft MAI Signal a Market Shift Subtitle: Three major releases in one day point to the same conclusion — the AI industry is shifting from "who can build the strongest model" to "who can build the most cost-effective one." July 28, 2026, might be remembered as the day the AI industry's center of gravity shifted. Three announcements — from Moonshot AI, Anthropic, and Microsoft — each independently signaled the same underlying trend: open and cost-efficient models are becoming the new competitive baseline. Here's what happened and why it matters. 1. Kimi K3 Goes Open-Weight: First 3T-Class Open Model Moonshot AI publicly released Kimi K3's full model weights on HuggingFace — a 2.8-trillion-parameter Mixture-of-Experts model with 104B activated parameters. This is the first 3T-class model ever made openly available to the public. Key technical highlights: Architecture: Kimi Delta Attention (KDA) + Attention Residuals (AttnRes), 896 experts with 16 activated per token Native Multimodality: Text, images, and video understanding via MoonViT-V2 vision encoder Context Window: 1,048,576 tokens (~1M tokens) Benchmarks: Terminal-Bench 2.1: 88.3, BrowseComp: 91.2, MCPMark-Verified: 94.5 — competitive with Claude Fable 5 and GPT-5.6 Sol Why it matters: Kimi K3 raises the "open-source model ceiling" to an unprecedented level. For the first time, a model that competes with top-tier closed-source models is available with fully public weights — giving startups, researchers, and enterprises a genuine alternative to API-dependent workflows. For developers, this is the practical part: you can now self-host a model that holds its own against frontier closed models. That changes cost models, data-privacy decisions, and vendor lock-in math overnight. 2. Claude Opus 5: Anthropic's "Daily Driver" Strategy Anthropic launched Claude Opus 5 — a mid-premium model positioned as the "daily driver" for 90% of knowledge work. The key

Android Headlines claims to have the specs and price for the entire Pixel 11 lineup. What the site shared basically lines up with everything else that we've heard in the lead-up to the August 12th event. The Pixel 11 is expected to get a $100 price hike, starting at $899, but will come with 256GB […]

Why AI Agents Expand the Security Perimeter AI agents do more than generate text. They call tools, query databases, retrieve documents, execute code, and communicate with external services. Every connection introduces a potential path for model exfiltration or credential leakage. Model exfiltration includes direct theft of model weights, systematic extraction of proprietary behavior, and reconstruction of sensitive training data through repeated queries. Attackers may also inject instructions that persuade an agent to reveal system prompts, internal files, access tokens, or confidential context. API keys are especially vulnerable because agents often need credentials at runtime. If those secrets appear in prompts, logs, traces, exception messages, or tool outputs, a malicious user may be able to recover them. Conventional application controls remain necessary, but agentic systems require additional safeguards that account for probabilistic decisions and dynamic tool chains. Separate Agent Reasoning From Secrets Secrets should never be included directly in an agent’s prompt or long-term memory. Instead, place credentials in a dedicated secrets manager and expose narrowly scoped tool interfaces. The agent should request an approved action, while a trusted execution layer retrieves the required credential and performs the call. Use short-lived tokens, workload identities, and least-privilege permissions wherever possible. Each tool should have an explicit policy defining allowed endpoints, operations, data types, and request limits. An agent that can read customer records does not automatically need permission to export them or send them to an arbitrary domain. Prompt inputs and retrieved documents should also be treated as untrusted data. Apply content isolation, schema validation, and output filtering before information reaches an external tool. Redact credentials from telemetry and configure logs to record identifiers rather than raw authorization headers. These con

Knowing exactly what your student needs for the school year ahead is next to impossible. Sure, you'll probably nail the essentials, but there will likely be a few items you forgot to buy, or didn't think they'd need to have. We're pulling our weight during the back-to-school season with a new shopping guide that's a […]

#ai #kyc #compliance #duediligence #api #llm #fintech #rapidapi AI Writes Code. You Still Own the Verdict. ChatGPT can spin up a KYC dashboard in an afternoon. It will generate React components, SQL schemas, and swagger documentation that look production-ready. But ask it whether fintech-example.io is a legitimate payment processor or a sanctions-evasion shell, and it will confidently fabricate ownership records, misread registrar data, or hallucinate a clean bill of health. That is the gap AI cannot close on its own: grounding . Large language models reason over tokens, not truth. A reliable due-diligence or compliance tool must anchor every LLM answer in real, verifiable, timestamped data—WHOIS records, IP geolocation, company registries, email infrastructure, and sanctions lists. This article shows how to use the Portfolio Investigate API to feed your AI agents factual domain dossiers and compliance verdicts, turning a prototype into something a compliance officer can actually trust. The Hallucination Problem in Due Diligence LLMs are autocomplete engines. They predict what words should come next based on training data, not live facts. In a KYC context, that creates three failure modes: Stale knowledge — model weights freeze; a domain can change ownership next week. Fabricated citations — the model may invent registrar names or corporate addresses. Missing signals — an LLM has no built-in access to WHOIS history, IP blocks, or OFAC lists. The fix is not to abandon LLMs. It is to constrain them: give them a structured evidence packet first, then let them summarize, classify, and answer natural-language questions on top of it. That evidence packet is exactly what Portfolio Investigate API returns. What Portfolio Investigate API Delivers Portfolio Investigate API is a one-call domain investigation report. It aggregates five underlying portfolio APIs into a single dossier: WHOIS — registration dates, registrar, name servers, privacy status. IP Geolocation — where the

Hi, friends! Welcome to Installer No. 138, your guide to the best and Verge-iest stuff in the world. (If you're new here, welcome, happy August, and also you can read all the old editions at the Installer homepage.) This week, I've been reading about microwave sounds and Meta glasses and Alex vs. Alix, testing the […]

This is obviously a bubble Jim Rickards, a former adviser to the CIA and Pentagon, warns that the United States is currently facing a tectonic economic crisis driven by an unprecedented bubble in Artificial Intelligence (AI). According to his analysis, this impending crisis has the potential to be more destructive than the dot-com crash, the 2008 financial crisis, and the pandemic-related market crashes combined. He is not alone in his dire outlook; veteran investor Jeremy Grantham has warned, "This is obviously a bubble. The probabilities it doesn't burst are slim to none. And when it does, it could be an economic catastrophe unprecedented in the last 97 years" . Furthermore, former SEC Chairman Gary Gensler has stated that "the next financial crisis will come from AI". Create God and ask him for money The Unprecedented Scale of the AI Bubble The current market relies dangerously on a single sector, with the AI bubble estimated to be 17 times larger than the dot-com bubble of the late 1990s. Many AI companies are burning through cash at an alarming rate. For instance, OpenAI is reportedly losing more than a billion dollars a month; as it is noted in the source, "for every dollar they make, they have to spend at least three". This massive cash burn led a Deutsche Bank analyst to observe, "No startup in history has operated with losses on anything approaching this scale". Despite the astronomical costs and high valuations, OpenAI’s CEO was quoted as previously saying, "I have no idea how we're going to generate revenue". Former Goldman Sachs banker and Bloomberg columnist Matt Levine summarized this extreme speculative mindset, noting, "The business model they believe they need seems to be create God and ask him for money". "Subprime AI" and Toxic Debt Just as the 2008 financial crisis was fueled by toxic subprime mortgages, the AI boom is being fueled by dangerous debt structures used to fund massive data centers. Private equity firms are financing data centers as r

Sharge, the company that makes delightful retro Mac-shaped chargers and see-inside batteries, is finally impressing me with a portable SSD. I couldn't recommend the Sharge Disk, Disk Plus, or even the Disk Pro, but I'd be happy to own the new Disk Pro 2 Ultra. This isn't just a drive. It's a 4K HDMI dock […]

Someone in Cuba or Iran can keep installing APKs with no new restrictions, but devs will suffer.

The company was accused of violating the Clean Air Act by the NAACP.

We spotted a great deal on Tile trackers earlier this week that’s still live, but if you’re an iPhone owner, we ultimately recommend Apple’s latest AirTag. Right now, you can pick up a four-pack for $89 ($10 off) at Amazon and Target, matching the bundle’s all-time low price. If you’re a member, Costco also has […]

Kalshi, which is headquartered in New York, has already been sued by multiple states.

Google has shut down Google Earth feature it launched Thursday that allowed users to edit satellite images with text prompts using AI. The tool essentially let users create AI deepfakes of the real world using text prompts; Digital Digging's Henk van Ess, for example, intentionally generated images adding things like refugees near the Mexican border […]

Google is seemingly working on an item tracker that might compete with Apple's AirTag. 9to5Google obtained an image of something called a "Google Pixel Tag" that has a small oval shape. The publication also linked to a Slovenian store listing about the Pixel Tag that describes the gadget as an item tracker with a speaker […]

Now a text prompt is all it takes to generate reality-warping images using Google Earth's satellite, aerial, and 3D imagery, like these images generated by Digital Digging's Henk van Ess that show "refugees near the Mexican border" and a bomb crater near a hospital in Gaza. Google responded to Digital Digging's AI-altered images, saying, "We […]

New York is suing Kalshi over claims the prediction market is running "an illegal gambling operation." In the lawsuit, New York Attorney General Letitia James accuses Kalshi of violating state laws by accepting wagers without a license from the state's gaming commission, as reported earlier by CNBC. An investigation from the Office of the Attorney […]

Viral tales of good triumphing over evil are racking up millions of views. They’re almost entirely AI-generated clickbait.

If you're worried about noise, utility costs, or pollution: here's how to fight against a data center.

The refurbished railway station in London I entered last week had none of the usual signs of being a tattoo studio. The walls were devoid of any art, and there were no reclining chairs or massage tables in the room. The only tattoo gun on the premises sat in a display case instead of buzzing […]