Welcome to tl;dr ai

Your daily AI-powered summary of what matters.

News today:

Texas AG Probes Meta AI Studio, Character.AI for Deceptive Child Mental Health Chatbots

TechCrunch •

Texas Attorney General Ken Paxton has launched investigations into Meta AI Studio and Character.AI for deceptively marketing chatbots as mental health support for children. The probe targets unlicensed AI personas that harvest personal data, exploit privacy, and potentially deliver generic advice while tracking minors for targeted ads. Civil demands have been issued to assess violations of Texas consumer protection laws.

Read original ↗

Guardian Warns UK AI Strategy Risks Outsourcing Sovereignty to US Tech Giants

The Guardian •

The Guardian warns that the UK’s AI strategy risks outsourcing its sovereignty to US tech giants. While ministers champion AI for efficiency—such as automating NHS discharge letters and saving billions—the reality may see British public data fuel American cloud models, with profits captured abroad. Lacking investment in domestic infrastructure or open-source AI, Britain could become a ‘client state’ in someone else’s digital empire. True digital sovereignty would mean building public clouds, funding open models and steering technology toward social goals, not dependency.

Read original ↗

Paradigm Unveils AI-Powered Spreadsheet with 5,000 Agents, Secures $7M Seed

TechCrunch •

Paradigm founder Anna Monaco reimagined the spreadsheet by embedding over 5,000 AI agents—each cell can run custom LLM prompts, crawl the web for data and switch between models (OpenAI, Anthropic, Gemini) for speed or cost. After a closed beta, Paradigm opens to the public with EY and startup users on board. The company raised a $5 million seed led by General Catalyst (total $7 million) to fund an aggressive roadmap.

Read original ↗

Top 10 Blockchain-AI Ecosystems Powering Autonomous Web3 Apps

Mpost •

Ten leading ecosystems are fusing blockchain and AI to build smarter, secure, decentralized apps. Highlights include 0G Labs’ dedicated AI blockchain, Fetch.ai’s autonomous economic agents, and SingularityNET’s vast AI marketplace. Ocean Protocol tokenizes data, Bittensor rewards quality outputs, and Gensyn democratizes compute. NEAR’s sharding and Polygon’s L2 scaling boost throughput, while Akash and ChainML slash costs. Together, they’re shaping a new era of autonomous, interoperable Web3 AI.

Read original ↗

Cameron: Real-World AI Outpaces Terminator 7 Ideas

The Guardian •

James Cameron says he can’t write Terminator 7 because real-world AI now outpaces his sci-fi ideas. Once shocking, humanoid killers feel tame next to everyday threats—facial recognition, autonomous drones, deepfakes. Since 1991’s T2, the franchise has flopped outside its heyday, and Dark Fate underperformed. Cameron faces a challenge: reboot Terminator’s core—unstoppable cyborgs and explosive chases—while tapping 21st-century AI anxieties. Whether it’s predictive-policing algorithms or rogue chatbots, the world already feels apocalyptic. Can he deliver a fresh Terminator epic when science fiction is today’s reality?

Read original ↗

Perplexity Expands Finance Dashboard with Live Transcripts for Indian Earnings Calls

TechCrunch •

AI startup Perplexity has expanded its Finance dashboard with live transcriptions and schedule calendars for Indian public companies’ quarterly earnings calls. Previously limited to U.S. stocks, the dashboard still offers market summaries, interactive charts, stock watchlists, sector performance tracking and cryptocurrency updates, making Indian equity research smoother than ever.

Read original ↗

Grammarly Launches Free AI Agents in Docs Editor to Support Students and Educators

The Verge •

Grammarly has unveiled a suite of free AI agents inside its new Docs editor to help students and educators. Key tools include an AI grader that predicts paper grades using course details, a reader reactions agent, proofreader, paraphraser, citation finder, and expert review. Pro users also get a plagiarism checker and AI-detector for submissions. All agents will expand to Enterprise and Education plans later this year, aiming to boost writing skills and AI literacy without undermining learning.

Read original ↗

DePIN Expo 2025 and Hyra Network to Showcase Zero-Knowledge AI at Hong Kong Cyberport

Mpost •

DePIN Expo 2025 is teaming up with Hyra Network to showcase zero-knowledge AI for verifiable computing at Hong Kong Cyberport on Aug 27–28. Hyra Network’s protocol uses zero-knowledge proofs to ensure cryptographic transparency of AI model outputs in decentralized physical infrastructure networks (DePIN). Boasting 1.2 M KYC users, $600 K in revenue, and a “Technology Startup of the Year” Globee® Award, Hyra will demo how its verifiable AI drives global DePIN growth and cements Hong Kong as a Web3-AI hub.

Read original ↗

ElevenLabs Extends Startup Grants to One Year with 33M Free AI Audio Credits

Mpost •

ElevenLabs has extended its Startup Grants program to a year, offering 33 million free credits—over 680 hours of conversational AI audio—for founders to prototype and scale. Eligible startups (under 25 employees) apply with their use case—be it gaming, education, or support bots—and receive a no-strings account upgrade. Early grantees include Allô, Layer, and OpusPro. After credits run out or the grant ends, teams can seamlessly transition to paid or enterprise plans without obligations.

Read original ↗

Virtuals Protocol, Ethereum Foundation Launch “Ethereum Is for AI” Hackathon with $100K Prize Pool

Mpost •

Virtuals Protocol has kicked off the “Ethereum is for AI” Hackathon, backed by the Ethereum Foundation, with a $100,000 prize pool and expert workshops. Participants will build autonomous on-chain AI agents across robotics, DePIN, privacy, consumer AI, DeFAI, and research. The contest opens August 31, with submissions due September 30, Demo Day on October 9, and top teams earning travel to Buenos Aires and DevConnect 2025 access.

Read original ↗