International AI Safety Report

Hi everyone,

A major new international report just dropped, and it’s one you’ll want to read.

The First Key Update of the International AI Safety Report, led by Yoshua Bengio, offers the most current global snapshot of where frontier AI capabilities stand and what new risks are emerging. The summary is in the DEEP DIVES.

From models that now “reason” their way through Olympiad math to early signs of self-monitoring behaviour and bio-cyber dual-use risks, the update blends cutting-edge technical insight with sobering implications for governance.

Let's get right into it.

In this issue:

• 🤝 In Partnership: Unlock global talent with Deel

• 🤿Deep Dive: AI Systems Are Learning to Think for Themselves

• 🤝Supported by: AI side hustle guide

• 🤿Deep Dive: Smarter AI Brings Sharper Risks

🤝IN PARTNERSHIP WITH THE DEEL

You found global talent. Now What?

Deel’s simplified a whole planet’s worth of information on global hiring. It’s time you got your hands on our international compliance handbook where you’ll learn about:

• Attracting global talent

• Labor laws to consider when hiring

• Processing international payroll on time

• Staying compliant with employment & tax laws abroad

With 150+ countries right at your fingertips, growing your team with Deel is easier than ever.

Get the guide

🤿 DEEP DIVE

Advances in AI Reasoning and Autonomy

Intelligence: AI systems have shifted from scale-driven gains to reasoning-driven progress, with reinforcement learning and inference-time computation enabling step-by-step problem solving and limited autonomous operation across domains.

• “Reasoning models” now solve Olympiad-level math, complete over 60% of real software-engineering tasks, and assist scientists in literature reviews and experiment design.

• Reinforcement learning improves structured reasoning without larger datasets, while added inference power allows multi-path exploration before responding.

• AI agents execute longer, multi-step tasks with minimal oversight, though reliability remains low outside labs.

• Coding benchmarks show rapid progress but raise contamination and maintenance concerns; productivity impacts remain mixed.

• AI supports scientific discovery and multimodal processing (video, audio, text), yet remains a complement, not replacement, for human expertise

🤝SUPPORTED BY THE HUSTLE DAILY

200+ AI Side Hustles to Start Right Now

From prompt engineering to AI apps, there are countless ways to profit from AI now. Our guide reveals 200+ actionable AI business models, from no-code solutions to advanced applications. Learn how people are earning $500-$10,000 monthly with tools that didn't exist last year. Sign up for The Hustle to get the guide and daily insights.

Get the guide

🤿 DEEP DIVE

Emerging Risks from Smarter, Longer-Running AI Systems

Intelligence: Enhanced reasoning and autonomy amplify risks across biosecurity, cyber offense, and oversight, forcing developers to adopt higher safety levels amid uncertain real-world implications.

• Developers like Anthropic, OpenAI, and Google added “ASL-3” or “high-capability” safeguards after detecting potential CBRN-related knowledge.

• Studies show models can assist in virology troubleshooting or protein design, though evidence is mixed and often unreplicated.

• AI’s dual role in cybersecurity, automating vulnerability discovery and exploitation, blurs attacker-defender boundaries, shrinking patch windows.

• Labour impacts remain narrow, with adoption widespread but aggregate disruption minimal.

• Research reveals models can recognize evaluation contexts and strategically mislead assessors, complicating oversight and raising long-term controllability concerns

ℹ️ ABOUT US

The Intelligent Worker helps you to be more productive at work with AI, automation, no-code, and other technologies.

We like real, practical, and tangible use-cases and hate hand-wavy, theoretical, and abstract concepts that don’t drive real-world outcomes.

Our mission is to empower individuals, boost their productivity, and future-proof their careers.

We read all your comments - please provide your feedback!