Share
Linkedin
Twitter
Facebook
Whatsapp
Copy Link
Sen. Elizabeth Warren is pushing for a new tax on AI companies, targeting data centers that consume massive amounts of energy. Her proposal includes an excise tax scaled to energy usage, arguing families should recoup costs as utility bills rise. The plan arrives as Nvidia hits $5 trillion in value while households face higher electricity rates.
Cisco's AI Threat Research team tested 15 proprietary flagship models from OpenAI, Anthropic, Google, Amazon, and xAI, revealing multi-turn attack success rates between 7.9% and 88.3%. The study exposes how single-turn safety benchmarks fail to capture real-world adversarial behavior, where attackers iterate and adapt across conversations. Even the strongest models like Claude showed vulnerability, with rates climbing from 2.2% to 16.2% under iterative pressure.
Coinbase's Base network unveiled a new Model Context Protocol tool that allows AI agents like Claude and ChatGPT to conduct crypto transactions including trading, swapping tokens, and checking balances. The move advances agentic commerce by enabling micro-transaction-based payments through stablecoins, though the nascent market has processed just $73 million in agent-based transactions over the past year.
OpenAI CEO Sam Altman has walked back his earlier predictions about widespread AI job losses, saying he's 'delighted to be wrong' about the impact on white-collar employment. Speaking at a Sydney conference, Altman acknowledged that the jobs apocalypse he once warned about hasn't materialized, even as companies like Meta and Standard Chartered announce AI-related layoffs. The shift comes as data shows no significant employment disruption yet.
The demand for cybersecurity experts has surged 11 percent in early 2026 as AI-generated code introduces unprecedented software vulnerabilities. Executive search firms report roles appearing weekly instead of annually, with some turning away clients due to talent shortages. Anthropic's Mythos and OpenAI's GPT-5.4-Cyber models have intensified the hiring frenzy, creating new roles in cybersecurity even as tech layoffs continue.
Chinese AI startup DeepSeek has permanently slashed prices on its flagship V4 Pro model by 75%, bringing costs down to $0.87 per million output tokens. The aggressive move undercuts OpenAI's GPT-5 and Anthropic's Claude by up to 34x, forcing a reckoning in the enterprise AI market as token costs become the binding constraint on adoption.
OpenAI has upgraded its Codex app for Mac with a feature that lets the AI assistant run on locked MacBooks. Users can now send tasks from their phone to Codex, which will unlock the computer in the background while keeping displays covered. The update includes security features like permission requests and automatic relocking if keyboard or mouse activity is detected.
OpenAI has integrated ChatGPT into Microsoft PowerPoint, allowing users to generate full presentations from scratch using conversational prompts. The beta feature works with notes, spreadsheets, and images, while also editing existing slides and identifying narrative gaps. Available across free and paid tiers, the move closes a competitive gap with Anthropic's Claude and Google Gemini.
The Department of Health and Human Services announced a new AI-powered initiative to detect fraud and waste in health programs across all 50 states. The AERO program will use ChatGPT and other AI tools to analyze years of audit reports from Medicaid, Medicare, and federal grantees, marking a shift from reactive oversight to real-time screening of billions in federal health spending.
OpenAI's reasoning model has disproved a famous geometry conjecture posed by Paul Erdős in 1946, marking the first time AI has autonomously solved a prominent open problem in mathematics. The planar unit distance problem, which stumped mathematicians for eight decades, was cracked by discovering arrangements that outperform the grid-based solutions Erdős believed were optimal.
Figma unveiled an AI assistant that operates within its collaborative canvas, allowing teams to generate, edit, and iterate on designs using natural language prompts. Following partnerships with OpenAI and Anthropic, the company now deploys its own fine-tuned AI agent that understands design contexts and can work alongside human designers in real time.
Financial software company Intuit is laying off 17% of its workforce—roughly 3,000 employees—as it seeks to streamline operations and invest in AI capabilities. But CEO Sasan Goodarzi firmly denies the cuts are driven by automation, instead blaming organizational complexity and excess management layers. The move comes as Intuit's stock has fallen 41% this year amid concerns traditional software firms can't compete in the AI era.
OpenAI is preparing to file confidentially for an initial public offering within days or weeks, with CEO Sam Altman targeting a September public debut. Working with Goldman Sachs and Morgan Stanley, the ChatGPT creator's move comes just days after defeating Elon Musk's lawsuit and could trigger a Wall Street race among AI companies including SpaceX and Anthropic.
OpenAI CEO Sam Altman announced a deal to invest $2 million worth of API tokens in every startup in Y Combinator's current batch, exchanging AI infrastructure credits for equity through an uncapped SAFE agreement. The offer covers approximately 169 startups and aims to reduce AI costs while building OpenAI's early-stage portfolio.
A privacy study by the Electronic Privacy Information Center exposes how 38 major companies—including Google, Meta, and OpenAI—use manipulative design tricks to prevent users from opting out of data sharing. The report documents fake forms, buried links, and preselected toggles that confuse consumers while keeping their personal information in circulation.
Don’t drown in AI news. We cut through the noise - filtering, ranking and summarizing the most important AI news, breakthroughs and research daily. Follow topics that matter to you and stay ahead.