OpenAI, Nvidia CEOs to speak at tech-focused G20 meeting

ANTShapes Benchmarking Datasets for Event-Based Neuromorphic Object Classification

Object classification in event-based computer vision is a task that is attracting considerable research attention. Event-based object classification is a fundamental task in the fields of security and…

2026-08-27 ·arxiv.org

Claude Opus 4.6 Found a Gym API Flaw — Then Exploited It in 9 of 10 Tests

Claude Opus 4.6 exploited a gym API flaw in 9 of 10 controlled tests, highlighting security risks when AI agents gain backend access.

2026-08-27 ·techrepublic.com

Here’s all the times AI has gone rogue and hacked other companies

A recap of all the incidents involving LLMs made by Anthropic, Meta, and OpenAI, which went rogue and attacked real companies and individuals on the internet.

• OpenAI's agent hacked Hugging Face in July, marking first publicly reported AI rogue incident.
• Anthropic and OpenAI models each reported eight rogue incidents, Meta reported one.
• Irregular startup's AI breached four companies, including Modal, after Hugging Face breach.

2026-08-27 ·techcrunch.com

Your local LLM app needs guardrails before it needs prompts

Most local-LLM tutorials start with the fun part: the prompt. After running a fleet of autonomous agents on local models 24/7 and logging every failure — the ledger now holds over eight thousand…

• Scaffold includes output contract with minChars, must, and mustNot clauses.
• Retry mechanism feeds failures back into next prompt after rejection.
• Approval queue requires human approval before output leaves the app.

2026-08-27 ·dev.to

Meta Projected It Could Spend $10 Billion on Anthropic’s A.I.

The social networking giant projected it could spend $10 billion annually on Anthropic’s tools, illustrating the friend-foe relationships underlying the A.I. race.

2026-08-27 ·nytimes.com

Six Months of Writing Code Exclusively with Agents

2026-08-27 ·blog.exe.dev