The Register AI

Howler launches ephemeral AI chat interface with automatic conversation deletion.

AI smartphone automation tool struggles with basic tasks while incurring token costs.

Open source maintainers report 25-31% of PRs are AI-generated low-quality submissions.

GNOME and System76 debate accepting AI-generated bug reports vs. requiring human contributions.

MeetTwins enables AI avatars to replace users in Google Meet calls.

KDE Akademy proposes AI-native desktop environment personalizing Plasma around user models.

Oracle claims AI improves SaaS installations and drives IaaS adoption against market decline.

Microsoft releases AI code migration tool for Salesforce and ERP users.

Microsoft Edge team automates quality checks after influx of AI-generated browser extensions.

AI mushroom identification models achieve only 65% accuracy, posing safety risks.

OpenClaw 2.0 improves UX but doesn't address underlying security issues in popular agent framework.

Microsoft postpones Teams Facilitator AI bot launch by 2 months for refinement.

Salesforce reports 50% of bookings driven by AI consumption via Flex Credits, raising monetization transparency concerns.

McKinsey finds enterprise AI spending increasing but ROI gains remain unclear and unproven.

Canonical funds Bristol researchers to explore AI-assisted C-to-Rust translation for legacy code modernization.

Generic commentary on software quality without specific AI relevance.

Slack Code integrates AI agents directly into group chat for collaborative development workflows.

Firefox iOS adds experimental ad blocking feature unrelated to AI.

Microsoft removes Excel Copilot side-pane function as users consolidate to simpler interface.

Microsoft delays Exchange subscription update due to AI-generated bug backlog management challenges.

OpenWALDO open-weight project seeks contributors to build 167B token transparent alternative to proprietary AI giants.

OpenAI launches $125/month premium tier amid competitive pressure from improving open-weight models.

NEC parking technology, Infosys Crocs deal, Fujifilm printer exit, Chinese datacenter developments.

Study finds human reviewers miss ~33% of dangerous requests in AI coding agent systems like Claude Code.

UK explores regulatory requirements for employer disclosure before deploying workplace surveillance AI tools.

ShieldFont open-source tool poisons font data to defend against unauthorized AI model scraping.

Microsoft reports strong cloud revenue growth but modest M365 AI monetization despite high capex.

Perplexity launches Model Council feature running up to 8 AI models simultaneously for comparative analysis.

Professor uses prompt manipulation to detect student AI cheating, confirming predictable human behavior patterns.

Opinion: Major AI models including Grok exhibit leftist bias in outputs.

ChatGPT Business/Enterprise users lack native export; third-party tools like scrapemychats fill gap.

IBM claims AI did not kill software deals, only delayed them due to enterprise hardware budget competition.

EU regulators force Google to share AI and search capabilities with competitors under digital competition rules.

Forrester warns AI infrastructure costs and usage-based pricing will significantly increase enterprise software budgets in 2025.

Rising AI infrastructure costs and token pricing are causing executives to reassess AI investment priorities and ROI assumptions.

Python tool automates job search, resume generation, and cover letter writing from web-scraped openings.

Startup offers tool to obfuscate AI-generated academic papers as human-written despite stated non-cheating claims.

CLI tool called Atrophy CLI helps developers maintain coding skills to counter skill degradation from AI-assisted coding.

Godot project bans AI-assisted code contributions, citing maintainer concerns about contributor understanding and code quality.

Microsoft adds bot authentication/filtering to Teams meetings, allowing ISVs to pre-authorize trusted bot integrations.

Anthropic ships Claude as always-on Slack agent, shifting from chat interface to proactive workplace automation.

Apple's redesigned Siri with on-device AI degrades user experience, bundling intelligence features that complicate basic OS search.

Microsoft positions Copilot as autonomous agent to manage workflows, raising trust and oversight concerns for always-on deployment.

Anthropic files for IPO after surpassing OpenAI's private valuation, marking major capitalization milestone for frontier AI startup.

Commentary on Google's integration of AI features across search and products degrading content quality and user experience.

Ontario audit reveals 60% of AI medical scribe systems misrecord drug prescriptions, exposing safety failures in clinical deployment.

OpenAI exec indicates company plans $50B annual compute spend to scale training, highlighting capital intensity of frontier model development.

Astera offers high-speed networking infrastructure alternative to NVLink for AI cluster connectivity.

Anthropic releases finance-focused agents enabling Claude to execute autonomous transactions and financial operations.

IBM, Google, and Intel launch AI automation agents for Db2 database administration, enabling autonomous DBA operations.