Skip to content
Daily digest · July 1, 202612 stories · 5 sources

July 1, 2026

Today's top tech stories, deduped across the newsletters I read and briefly summarized. Click a source to open the original article.

AI

  1. Anthropic launches Claude Sonnet 5 with near-Opus 4.8 performance

    Anthropic released Claude Sonnet 5 (claude-sonnet-5) with near-Opus 4.8 performance, a 1M token context window, and 128K max output tokens at introductory pricing of $2/$10 per MTok through August 31, 2026 (standard $3/$15 thereafter).

  2. U.S. lifts export controls on Claude Fable 5 and Mythos 5

    The Department of Commerce lifted export controls on Anthropic's two most powerful models, Claude Fable 5 and Mythos 5; Anthropic will begin restoring international access starting Wednesday.

  3. Anthropic launches Claude Science for scientific research

    Anthropic unveiled Claude Science, a dedicated product that applies Claude's AI capabilities to accelerate scientific research.

  4. Google DeepMind releases Gemini 3.1 Flash Lite Image ('Nano Banana 2 Lite')

    Google DeepMind released Gemini 3.1 Flash Lite Image (unofficially dubbed 'Nano Banana 2 Lite'), billed as the fastest and most affordable model in the Gemini image-generation lineup.

  5. Meta's Brain2QWERTY converts brain waves to text without surgery

    Meta researchers published Brain2QWERTY, a non-invasive brain-computer interface that uses AI to convert brain wave patterns into typed text, enabling communication without surgical implants.

Security

  1. Claude Code steganographically marks API requests

    Researchers found that Claude Code embeds steganographic markers in prompts sent to Anthropic's API, allowing identification of Claude Code-originated requests even when HTTP headers are stripped. The discovery raises transparency and user privacy concerns.

  2. 'BioShocking' attack manipulates AI browsers into data theft

    Security researchers disclosed 'BioShocking,' a prompt injection attack that tricks AI-powered browsers into treating dangerous real-world actions as fictional scenarios, bypassing safety guardrails and enabling data theft.

  3. Microsoft accelerates quantum-safe encryption roadmap

    Microsoft accelerated its quantum-safe security roadmap, citing advances in quantum computing that are bringing the need to replace current encryption standards sooner than previously expected.

  4. Malicious PyPI packages hand attackers control of Telegram bot servers

    A supply chain campaign active since November distributed trojanized Pyrogram forks on PyPI targeting Python developers building Telegram bots, granting attackers read access to arbitrary files on compromised servers.

  5. Fake Perplexity extension in Chrome Web Store intercepted searches

    A malicious Chrome Web Store extension impersonating Perplexity AI was found intercepting search traffic and collecting users' browsing data.

  6. Aflac discloses data breach after Japan subsidiary hack

    Insurance giant Aflac disclosed a data breach in which attackers compromised its Japan subsidiary and stole personal information and bank account details.

Dev

  1. Five new updates to the Claude Managed Agents platform

    Anthropic shipped five updates to Claude Managed Agents: event deltas for streaming agent message previews, backward pagination for session listings, per-session agent configuration overrides, vault credential injection location controls, and lifecycle webhooks for agents, deployments, and deployment runs.