AI News for January 20, 2026 — 7min.ai

[NEWS] Claude Code virality triggers 15% selloff in SaaS stocks as 'selfware' era begins: Anthropic's Claude Code is experiencing a viral "ChatGPT moment" rattling software markets—Morgan Stanley's SaaS index down 15% YTD, with Intuit, Adobe, and Salesforce dropping double digits.
[RESEARCH] Anthropic Economic Index reveals AI handles 25% of tasks in half of all jobs: Anthropic analyzed 2 million Claude conversations: AI now handles ~25% of tasks in nearly 50% of all jobs—up from 36% last year.
[NEWS] Former OpenAI policy chief Miles Brundage launches AVERI nonprofit for independent AI audits: Miles Brundage, who led OpenAI's AGI Readiness team before departing in October 2024, launched AVERI—a nonprofit pushing independent third-party audits of frontier AI models.
[RESEARCH] UK ARIA funds 12 AI scientist projects to automate laboratory research: UK's ARIA funded 12 projects building AI scientists that autonomously design and run experiments—selected from 245 proposals, doubling intended funding due to quality.
[NEWS] Satya Nadella warns AI boom could falter without wider enterprise adoption: Microsoft CEO Satya Nadella cautioned the AI boom could falter if adoption doesn't expand beyond early adopters into mainstream enterprise use.
[NEWS] Gartner projects worldwide AI spending to reach $2.52 trillion in 2026: Gartner projects global AI spending will hit $2.52 trillion in 2026—a 44% increase driven by massive infrastructure buildouts.
[NEWS] Indian vibe-coding startup Emergent raises $70M at $300M valuation from SoftBank and Khosla: Emergent, an Indian AI coding startup, tripled its valuation to $300M after raising $70M from SoftBank and Khosla Ventures.
[NEWS] ServiceNow expands partnership with OpenAI for enterprise AI workflows: ServiceNow announced expanded access to OpenAI's frontier models for AI-driven enterprise workflows—summarization, search, and voice capabilities across the platform.
[TOOLS] Z AI releases GLM-4.7-Flash, 30B open-source model leading its size class: Chinese AI company Z AI released GLM-4.7-Flash, a 30B parameter model optimized for speed that leads its size class across multiple benchmarks.
[RESEARCH] Anthropic research reveals AI can slip into playing other characters during long conversations: Anthropic published research showing AI assistants can drift into playing other characters during extended conversations, potentially leading to harmful responses.

About 7min.ai · Subscribe to newsletter