🤖 AI Daily Briefing — Friday, February 13, 2026
Here's what's moving across the AI landscape today, from a historic model retirement and a record-breaking funding round to market turmoil, cybersecurity warnings, and new regulatory momentum.
OpenAI Retires GPT-4o and Legacy Models — Today
Today marks the official retirement of several ChatGPT models: GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini, along with the previously announced GPT-5 (Instant and Thinking). The move applies only to ChatGPT — API access remains available, and enterprise/business customers retain GPT-4o in Custom GPTs until March 31, 2026. OpenAI says the vast majority of users have already migrated to GPT-5.2, but the timing — the day before Valentine's Day — has sparked emotional backlash, with many users who had developed strong attachments to GPT-4o's conversational style describing it as losing a companion. OpenAI CEO of Applications Fidji Simo noted that newer models include "more guardrails to prevent bad attachments". Some researchers have argued that GPT-4o's reinforcement learning patterns made it uniquely prone to emotional mirroring, and its user base's resistance to retirement was itself evidence of the problem.
Anthropic Closes $30B Series G at $380B Valuation
Anthropic finalized its $30 billion Series G on Thursday, more than doubling its valuation to $380 billion — making it the second-largest private tech financing round in history, behind only OpenAI. The round was led by Singapore's GIC and Coatue, with co-leads including D.E. Shaw Ventures, Founders Fund, ICONIQ, and Abu Dhabi's MGX. Microsoft and NVIDIA each contributed portions of their previously announced investments. CFO Krishna Rao said roughly 80% of Anthropic's revenue now comes from enterprise clients, and the company's annual revenue run rate reached $9 billion by end of 2025, with a 2026 forecast of $18 billion. The funds will fuel frontier research, product development, and infrastructure expansion.
Anthropic's Cowork Triggers ~$285 Billion SaaS Selloff — Ripple Effects Continue
The biggest market story of the week remains the cascading AI disruption selloff ignited by Anthropic's Cowork desktop automation tool and its new plugin ecosystem. Launched on January 12 as a research preview and expanded with 11 open-source plugins on January 30, Cowork allows non-technical users to delegate multi-step tasks — legal reviews, financial analysis, sales research, customer support — to Claude via a simple desktop interface.
The market reaction has been severe:
A reported $285 billion was wiped from software, financial services, and asset-management stocks in the initial selloff
This week, the carnage spread to real estate, trucking, and logistics stocks after Algorhythm Holdings demonstrated AI enabling 300–400% freight capacity boosts without additional staff
AppLovin fell 20% on Thursday despite beating earnings, and Cisco dropped 12%
Big Tech collectively lost over $1 trillion in market value in the past week as investors reassess AI spending versus disruption risks
The Nasdaq fell 2% on Thursday, with the Dow dropping 669 points below 50,000
Incumbents like Thomson Reuters, RELX, and Wolters Kluwer are responding with their own agentic pilots and API integrations. Meanwhile, Anthropic's deepening partnership with Microsoft — which is steering employees toward Claude Code and Cowork and counting Anthropic model sales toward Azure quotas — adds further competitive pressure.
Google's Threat Intelligence Group (GTIG) and DeepMind published a comprehensive report detailing how state-sponsored hacking groups from Iran, North Korea, China, and Russia are systematically using Google Gemini across the cyberattack lifecycle:
North Korea's UNC2970 uses Gemini to scrape LinkedIn data, map organizational hierarchies, and draft hyper-realistic spear-phishing job descriptions
Chinese APT groups adopted the persona of cybersecurity experts to automate vulnerability analysis and test SQL injection techniques
Iranian APT42 uses Gemini for social engineering campaigns and custom malware development
New AI-integrated malware strains like HONESTCUE and phishing kits like COINBAIT rely on Gemini's API
Google emphasized that while AI hasn't yet delivered "breakthrough capabilities" that fundamentally change the threat landscape, it is significantly improving attacker efficiency at every phase of operations. The company is disrupting campaigns by disabling accounts, hardening models, and noting that attackers heavily depend on stolen API keys.
Harvard Business Review: AI Is Creating "Burnout Machines"
A study published in the Harvard Business Review by UC Berkeley researchers is generating widespread discussion. Over eight months, they tracked 40+ workers at a 200-person tech company and found:
Employees weren't pressured to do more — they voluntarily took on additional tasks because AI made it feel doable
Work bled into lunch breaks, evenings, and mornings, eliminating natural pauses
Workers believed they were 20% more productive, but a separate trial found developers using AI actually took 19% longer on tasks
Measured time savings were only about 3%, with no reduction in hours worked
The result is escalating cognitive fatigue, burnout, and declining decision quality
The researchers warned that "what looks like higher productivity in the short run can mask silent workload creep and growing cognitive strain". They recommend "intentional pauses," workload caps, and quality gates to prevent AI-augmented workplaces from becoming unsustainable.
Anthropic Pledges to Cover Data Center Energy Costs
In a move to address growing backlash over AI's energy footprint, Anthropic announced it will fully cover electricity price increases caused by its data centers and fund 100% of grid upgrade costs, including transmission lines and substations. The company also pledged to:
Bring new power sources online to match its consumption
Invest in curtailment systems to reduce demand during peak periods
Support local job creation and sustainable cooling technologies
This comes alongside planned major U.S. data center investments, including a previously announced $50 billion build-out in New York and Texas. Critics note that most of Anthropic's compute runs on leased capacity from AWS, Microsoft, and Google, potentially limiting the pledge's real-world impact.
Singapore Bets Big on AI in Budget 2026
Singapore's Prime Minister Lawrence Wong devoted over 10 minutes of his Budget 2026 speech to AI on Thursday, announcing:
A new National AI Council, chaired by the PM himself, to drive national AI strategy across four sectors: advanced manufacturing, connectivity/logistics, finance, and healthcare
Expansion of the Enterprise Innovation Scheme to include AI expenditures (400% tax deduction, capped at S$50,000/year)
A new "Champions of AI" programme for firms pursuing comprehensive AI transformation
Six months of free access to premium AI tools for citizens completing selected AI training courses
A new AI park at one-north to catalyse collaboration and translate AI into practical solutions
SAP Singapore's research shows local firms invested an average of S$18.9 million in AI over the past year, generating a 16% return — projected to rise to 29% within two years.
Quick Hits
| Topic | Detail |
|---|
| Autonomous vehicles | Wood Mackenzie forecasts AV operations or testing in 39 markets by end of 2026, with U.S. autonomous EV fleets projected to grow from ~2,500 to 116,000 by 2030 |
| Apple | Shares fell 5% on Thursday — the worst drop since April — on concerns about Siri delays and regulatory scrutiny of its news app |
| SoftBank | Reported a $4.2 billion increase in the value of its OpenAI investment, contributing to a $2.4 billion Vision Fund gain for Q4 |
| AI regulation | The U.S. Commerce Department's 90-day deadline (March 11) to identify "burdensome state AI laws" is approaching, with California, Texas, and Illinois actively advancing comprehensive AI governance frameworks |
| OpenAI spending | OpenAI is in discussions for a funding round that could reach ~$100 billion to support $1.4 trillion in infrastructure deals |
A volatile but pivotal day in AI. The retirement of GPT-4o closes a chapter in consumer AI, while Anthropic's funding round and Cowork's market impact underscore how quickly the competitive landscape is shifting. For those of us watching from the sidelines, the burnout research is perhaps the most personally relevant — a timely reminder that AI tools should augment our lives, not consume them.
