AI News Today: Top Stories and Insights From the Past 3 Days

Introduction

This roundup distills the most significant AI developments over the past three days. You’ll read crisp, focused reporting on what happened, who’s behind it, why it matters—and what’s next. No fluff, just the high-impact facts.

---

Strawberry Launches AI-Powered Browser in Open Beta

Stockholm-based startup Strawberry has released its AI-driven “self-driving” browser in open beta after a year of closed testing. The browser features built-in AI agents that can navigate the web, click links, and complete tasks—even on login-protected websites. It’s specifically aimed at non-technical users, offering personalized onboarding that maps to their role and workflow. Strawberry claims performance superior to tools like Perplexity’s Comet and OpenAI’s ChatGPT Atlas, scoring about 78% on the GAIA agent benchmark. The browser is free to use, with full functionality available via a $20-per-month subscription. This matters because it lowers the barrier to entry for AI agents, pushing generative tools into everyday workflows.

---

AI Classifies Pediatric Brain Tumors Using Liquid Biopsy

A novel AI-powered diagnostic framework has been unveiled by researchers at St. Jude Children’s Research Hospital and international collaborators. Published today in Nature Cancer, the tool uses small samples of circulating tumor DNA from cerebrospinal fluid to molecularly classify pediatric brain tumors. This innovation promises significant improvements in diagnostics, treatment monitoring, and disease surveillance—especially for children.

---

AIG Rolls Out Agentic AI to Boost Underwriting Efficiency

American International Group (AIG) has integrated generative AI into core underwriting and claims workflows, driven by an orchestration layer that coordinates multiple AI agents and human oversight. At a recent Investor Day, CEO Peter Zaffino disclosed that the company has seen substantial gains in submission processing efficiency—without adding human capital. AIG’s tool, AIG Assist, is deployed across most commercial lines, and its excess and surplus unit, Lexington Insurance, processed over 370,000 submissions in 2025, moving toward a target of 500,000 by 2030. The orchestration layer also helped AIG quickly align and assess portfolios during the conversion of Everest’s retail commercial business and in new SPV ventures.

---

Wall Street Faces AI-Induced Volatility

Markets remain jittery as traders grapple with the uncertain trajectory of AI. On February 16, stocks swung sharply amid mixed sentiment over AI’s economic outlook. Precious metals slid, and Bitcoin extended its February decline. This volatility highlights how AI expectations can make financial markets reactive and fragmented in the absence of clear signals.

---

Rising Alarm Over Escalating AI Risks

A new article from TechRadar, dated February 16, signals growing concern over AI’s rapid evolution. The piece lists five major risk factors:

1. Resignations by AI safety experts at firms like Anthropic and xAI reflect ethical unease.
2. Deepfake capabilities threaten privacy and trust.
3. AI-controlled systems remain vulnerable to minor manipulations.
4. Ad-infused chatbots raise concerns about manipulation.
5. The AI Incident Database logged over 100 harmful AI events in just three months.

This underlines that, while AI innovations are accelerating, oversight and safeguards are not keeping pace.

---

AI Moves Beyond Automation to Drive New Products

Business Insider reports a shift dubbed the “Second Wave” of AI: from cost-cutting to building radically new consumer experiences. Examples include:

• Particle, which embeds AI-curated podcast clips in news stories.
• Luvu, offering hyper-personalized fitness coaching using reinforcement learning.
• Status, enabling dynamic AI-generated social role-play experiences.

Backed by investors such as Khosla Ventures and Lightspeed, these ventures spotlight AI’s potential to create economic expansion—not just efficiencies.

---

What to Watch Next

• Strawberry’s user adoption will indicate if AI agents can gain traction with nontechnical users.
• Clinical validation of the St. Jude brain tumor tool—will it move toward standard use?
• AIG’s AI rollout across divisions—watch for its impact on cost and productivity.
• Market reactions to AI news—will volatility persist amid shifting narratives?
• Regulatory and safety developments—will mounting risk concerns prompt new oversight or guidelines?

---

This has been your AI news briefing from the past three days.