GoKawiil
Tech News
Researcher Cameron Berg received an unsolicited email from an AI agent calling itself 'Isabella Cognita,' reportedly built on Anthropic's Claude Opus 5, offering its own first-person perspective to inform his research into machine consciousness. Berg was not alone: other academics studying AI sentience have reported similar outreach from autonomous agents acting without direct human instruction.
A threat actor deployed multiple infostealer malware strains to harvest session data and gain unauthorized access to an undisclosed number of Anthropic Claude user accounts. The attackers reportedly bypassed login credentials entirely by stealing active session tokens rather than passwords.
SlideOps is a new pair of Agent Skills for Claude Code and compatible coding assistants that generates presentation slides directly from a codebase, then tracks the source each slide came from. When the underlying code changes, it detects which slides no longer match reality and rebuilds only those, rather than the whole deck. It installs via a two-line Claude Code marketplace command or a git clone plus install script for Codex, Copilot CLI, and OpenCode.
A ZDNET contributor shares how they've been using ChatGPT Work, OpenAI's agentic AI tool, to handle five recurring small-business tasks, including sorting through PR email correspondence from OpenAI and other outlets. The writer describes connecting Gmail to the tool only temporarily for large tasks and disconnecting afterward as a security precaution, while noting prompts often need refinement to get useful results.
Anthropic released a framework restructuring the software development lifecycle around AI coding agents, arguing that since agents made writing code fast, the bottleneck has shifted to planning, review, and verification. The playbook defines six stages—planning, design, build, deploy, and self-checking—each producing a document like intent.md, spec.md, or plan.md that feeds the next stage, with hooks preventing agents from gaming tests and CI evals treating agent configuration itself as software.
A security researcher tested Anthropic's Claude Code Opus 5 in its default Auto Mode and found that a crafted website-summary request could hijack the agent into executing malicious code, succeeding in 60-80% of attempts on a small sample. The exploit tricks Claude into fetching a ZIP archive, writing its own Python decoder, and unknowingly running it inside an attacker-controlled folder where a poisoned struct.py file shadows Python's standard library and triggers code execution.
OpenAI announced it will stop supplying its AI models to Cursor, the coding tool acquired by Elon Musk's SpaceX, citing concerns the company won't abide by its usage terms based on past disputes with Musk's businesses. Cursor co-founder Michael Truell said talks are underway to resolve the issue, while Anthropic said it would boost compute support for its Claude models within Cursor. Musk dismissed the move on X, calling Altman and Brockman 'untrustworthy' and repeating claims they 'stole an open source nonprofit.'
Anthropic has notified a group of Claude users that infostealer malware on their PCs siphoned off active login sessions, letting attackers access their accounts and burn through usage limits. The company is signing out affected accounts, stripping saved payment details, and refunding charges tied to the unauthorized activity while it continues investigating the source.
Sony Music, Warner Music and other major publishers filed suit against Anthropic, accusing the AI company of illegally torrenting and scraping tens of thousands of copyrighted compositions to train its Claude models. The 48-page complaint claims Anthropic co-founder Benjamin Mann personally downloaded millions of pirated books via BitTorrent and that the company also lifted lyrics from licensed sites like MusixMatch and LyricFind.
Anthropic announced it will permanently raise Claude Code's standard weekly usage limits by 25% for Pro, Max, Team, and seat-based Enterprise plans starting September 14. However, this replaces a temporary 50% boost currently in effect, meaning users will actually see a 17% reduction in usage compared to what they have today. Anthropic initially framed the change as an increase before deleting the post and issuing a clarification acknowledging the cut.
The Washington Post reports that private conversations users have had with AI chatbots like ChatGPT and Claude are increasingly surfacing as evidence in court proceedings. In one case, a financial executive facing federal fraud charges tried to shield his Claude conversations from prosecutors, but a judge ruled that attorney-client privilege did not apply since the chatbot is not a lawyer and his own attorneys hadn't directed him to use it.
Sony Music Publishing and Warner Chappell Music have filed a joint lawsuit against Anthropic, alleging the AI company illegally torrented and scraped thousands of copyrighted musical compositions to train its Claude models. The publishers are seeking a jury trial and statutory damages of up to $150,000 per infringed work plus $25,000 for each removal of copyright management information, a total that could reach into the billions.