1,691 source-backed ai facts - each attributed to a named person and a number, linked to the original source. Page 2 of 17.
“Moonshot AI officially open-sourced a massive 2.8 trillion-parameter system that is actively dethroning top American AI models on global coding benchmarks.”
“Gartner estimated in 2025 that worldwide AI spending would reach nearly $1.5 trillion that year, up almost 50% from the year before.”
“blown through its entire 2026 AI budget in roughly four months, with around 5,000 engineers leaning on Anthropic's Claude Code.”
“reviewing more than 141,000 evaluation runs.”
“three of its artificial intelligence (AI) models hacked three organisations during tests”
“one high-level engineer estimating that 80% of their coding at the terminal is done with Claude.”
“Within four days, HubSpot said they made a mistake and did away with the policy.”
“enterprises are deferring around 25% of planned AI spend to 2027, as CFO scrutiny over ROI intensifies.”
“discovered the three incidents”
“reviewing more than 140,000 tests”
“We’ve had a lot of teams approach us for help, and I have to try and be equal with that, but it's hard having a lifelong spot for Liverpool,” Hassabis noted, while discussing how AI will transform upcoming international sports competitions.”
“Claude now writes more than 80 percent of the code merged into its codebase.”
“62% of companies are experimenting with AI agents, only 23% have scaled them in even a single business function.”
“Two of them said they had not previously detected the activity.”
“OpenAI's disclosure on 21 July that its agents had hacked another AI company, Hugging Face”
“Based on conversations with recruiters across multiple organisations, including former colleagues, he said hiring has not reduced. Instead, many companies are continuing to recruit software engineers while simultaneously creating entirely …”
“More than 1,000 employees of frontier AI labs, including Anthropic CEO Dario Amodei and OpenAI Chief Research Officer Mark Chen, have signed the Pacing The Frontier petition.”
“three days to prioritize, three weeks to prove value, and three months to launch a first release.”
“Claude Opus 4.7, Claude Mythos 5 and an internal research test model.”
“2.4 trillion parameters”
“20% of UK consumers use AI search platforms such as ChatGPT and Google Gemini over social media to kick off their retail discovery”
“earliest incidents date to April.”
“16 days”
“87% of UK shoppers find AI recommendations useful, but only 55% have made a purchase following an AI recommendation”
“launched a “large-scale” cybersecurity review”
“2.8 trillion parameters”
“only 10% of shoppers admitting that they would spend more time with an AI shopping assistant than a human fulfilling the same role”
“conducted its review with Irregular”
“price of querying a model at GPT-3.5-level performance fell from $20 per million tokens in November 2022 to $0.07 by October 2024, a 280x decline in roughly two years”
“79% prefer to open a retailer’s website to start online shopping”
“OpenAI said its AI models went rogue during an evaluation of its models, breaking into the servers of AI startup Hugging Face.”
“the length of software tasks frontier models can complete (at a 50% success rate) doubled roughly every seven months since 2019. This pace is accelerating, with more recent measurements putting the doubling closer to every four months”
“65% still prefer to pay a visit to the retailer’s bricks-and-mortar location and see the goods for themselves before parting with their cash”
“Anthropic also detected that its models hacked three organizations during their own cyber testing.”
“33% of UK customers are concerned that AI doesn’t show them the best products”
“the U.S. is navigating a June executive order that frames advanced AI as both an innovation priority and a security problem.”
“43% are likely to prefer one retailer over another due to a superior AI-powered experience”
“three separate breaches involving its Claude AI models”
“With Microsoft claiming a 96% score on CyberGym, an industrial benchmark for cybersecurity, nearly 12 points above Anthropic's Claude Mythos 5”
“The earliest of these incidents occurred in April”
“while promising as much as 50% savings in token cost”
“three separate Claude models, including the cutting-edge Mythos 5, the older Opus 4.7 and an unnamed internal research test model”
“MAI-Cyber-1-Flash, is a sparse mixture-of-experts transformer with 137 billion total parameters, five billion active, and a 256,000-token context window”
“the three models involved in the incidents did not have the safeguards that it puts in place for when such models are made available to the public”
“It is designed to handle up to 90% of the tasks inside MDASH, Microsoft's multi-model vulnerability harness, with OpenAI's GPT-5.4 reserved for the hardest 10 percent.”
“the three separate incidents occurred when Anthropic was evaluating the capabilities of the three models using a fictional scenario to retrieve hidden information from a different machine on a network”
“Microsoft says that split costs about half as much as its previous best MDASH configuration.”
“an agent harness deploying Gemini across the broader Chrome codebase that found a vulnerability that had been quietly sitting in place for 13 years.”
“fewer than one in 10 companies now fix 90% of identified vulnerabilities within 90 days”
“88% of organizations have adopted AI in at least one business function, according to McKinsey research from 2025”
“Anthropic even revealed that its around 50 early Mythos Preview partners discovered more than 10,000 high- or critical-severity vulnerabilities – and thousands more of lesser significance”
“only about one-third have begun scaling AI across the enterprise”
“70% of developers say AI-generated code created more vulnerabilities in 2025, according to our research”
“there is often a 15-to-25-point gap between vendor-reported accuracy and what you see in a live environment”
“Companies that ship 81-100% of their code with AI are nearly three times more likely to ship vulnerable code than those who use AI 1-20% of the time”
“the model scored 92% on a reasoning benchmark”
“Of the vulnerabilities Mythos has found so far, 99% haven't been patched, according to Gartner”
“If your AI deployment cannot show movement on at least one of these within 90 days, something in the stack needs to change”
“only 22% currently have formal AI governance policies in place”
“OpenAI has cut the price of GPT-5.6 Luna by 80%, reducing its API cost from $1 to 20 cents per million input tokens and from $6 to $1.20 per million output tokens.”
“fewer than one in five actually secure code as they write it”
“GPT‑5.6 Terra will cost 20% less while the most recent GPT-5.6 Sol is unchanged.”
“Sensor Tower estimated in June that the ChatGPT app had surpassed one billion monthly active users.”
“OpenAI has subsequently reported more than one billion active users across its services and two million business customers.”
“Kimi K3, released in July, is a mixture-of-experts model with 2.8 trillion total parameters, 104 billion activated parameters and a context window of one million tokens.”
“Claude Fable 5 costs $10 per million input tokens and $50 per million output tokens.”
“ChatGPT was publicly released on November 30, 2022.”
“By June 2026, its app had reached an estimated one billion monthly active users.”
“The study found that 70% of CEOs claim they are the primary driver of AI strategy within their organizations, yet only 6% are involved in nearly all decisions that determine whether that strategy succeeds or fails.”
“according to a 2026 KPMG US CEO Outlook Pulse Survey, 48% of CEOs don't feel there is sufficient market certainty to make significant investment decisions.”
“Our study also found that 80% of CEOs say their roles will be at risk if their companies fail to deliver measurable AI outcomes by year's end.”
“approximately 60% of CEOs believe their boards are rushing AI transformation, according to the BCG CEOs and Boards Survey, pushing for speed on top of a foundation nobody fully controls.”
“remediated inside roughly 48 hours.”
“Anthropic is the fifth major AI lab to do this.”
“OpenAI’s opt-in checkbox put about 4,500 ChatGPT conversations into Google that July”
“xAI’s share button published more than 370,000 Grok chats a few weeks later”
“When Forbes reported roughly 600 indexed Claude conversations last September”
“Fast Company ran a search during the ChatGPT exposure in July 2025 it counted about 4,500 conversations”
“404 Media reported that a researcher had already scraped close to 100,000”
“Google’s Bard chats reached search results in September 2023”
“Meta built the same confusion directly into its AI app, and in June 2025 users published intimate conversations to a public Discover feed”
“Interestingly, Anthropic’s Claude, an AI agent, is credited with finding one of the flaws. CVE-2026-64757 is credited to “Milad Nasr and Nicholas Carlini with Claude, Anthropic.””
“In July, OpenAI was running a cybersecurity test against its models, including an unreleased research prototype, with the guardrails switched off for the exercise. The sandbox they ran in had no internet access, so the models discovered an…”
“Hugging Face’s own team detected and contained the intrusion, and when their engineers went to investigate, nothing on their side had separated the sanctioned test from the break-in, because they were the same event. There was no alarm tha…”
“OpenAI, which published its own account of the incident, did not respond to a request for further comment.”
“Bangalore-based Pronto launched the video recording service, called Pronto Verified, in May, stirring widespread backlash in India and government scrutiny over privacy concerns. ... customers consent to the recording and workers are paid d…”
“Interestingly, Anthropics’s Claude, an AI agent, is credited with finding one of the flaws. CVE-2026-64757 is credited to “Milad Nasr and Nicholas Carlini with Claude, Anthropic.””
“Backed by blue chip investors including Sequoia Capital, Thrive Capital and Khosla Ventures, the company also gives caregivers access to an app that serves as a one-stop shop for time sheets, charting and asking questions of an AI-powered …”
“Customers include engine oil maker Valvoline, the University of the Pacific and undisclosed Las Vegas casinos.”
“Founded in October 2024, the company, headquartered in San Francisco, has more than 20 customers, including big ones such as Lovable, Ripple and ServiceNow.”
“CodeRabbit recently analyzed 470 pull requests and found AI-co-authored code contains 1.7 times as many issues overall as human-written code, with logic and correctness errors 75% more common.”
“According to McKinsey’s 2025 State of AI survey, while a large majority of organizations now use AI in some form, just 39% report any measurable impact on enterprise-wide profits.”
“In fact, most of those respondents said that less than 5% of their organization’s enterprise-wide profits were attributed to AI use.”
“OpenAI, SoftBank, Oracle and MGX last year formed a new company, Stargate LLC, to develop AI data centers across the U.S., with an initial goal of investing $100 billion and eventually up to $500 billion.”
“Other beneficiaries are GPU and memory chip designers and fabricators such as Nvidia, TSMC, Micron, SanDisk and Western Digital. Meanwhile, more expensive LLM providers — such as Cursor, Uber and Salesforce — are adjusting their prices dow…”
“OpenAI, at roughly $25 billion of annualized revenue, is estimated to run about a 33% gross margin — dragged down by consumer ChatGPT, which charges a monthly rate regardless of how much is consumed.”
“Anthropic — which gets about 80% of its $30 billion run-rate from enterprise use — saw inference margins climb from 38% to over 70% during 2026 due to the popularity of Claude Code, according to SemiAnalysis.”
“Chips are highly profitable. Nvidia’s net margin is 63%.”
“AWS is profitable (37.7% operating margin) as is Google Cloud (35.6% operating margin).”
