Can We Secure AI With Formal Methods?

Can We Secure AI With Formal Methods? November-December 2025

Give thanks for benchmarks with “Veri-” in the title.

Nov 29, 2025 • Quinn Dougherty

1

October 2025

July-October 2025 Progress in Guaranteed Safe AI

Hot FMxAI Summer grinds along. Many new founders on the scene. Thinking of rebranding the newsletter away from “guaranteed”.

Oct 22, 2025 • Quinn Dougherty

4

3

June 2025

May-June 2025 Progress in Guaranteed Safe AI

Benchmarks! Get your benchmarks! A startup and a nonprofit lab each launch

Jun 26, 2025 • Quinn Dougherty

1

April 2025

March-April 2025 Progress in Guaranteed Safe AI

Who even pays for patching open source software? And ARIA news

Apr 20, 2025 • Quinn Dougherty

February 2025

January-February 2025 Progress in Guaranteed Safe AI

Proving the Coding Interview, R1, a job, and yet another ARIA solicitation

Feb 28, 2025 • Quinn Dougherty

2

January 2025

November-December 2024 Progress in Guaranteed Safe AI

You missed my hot takes on FrontierMath, ho ho ho

Jan 22, 2025 • Quinn Dougherty

October 2024

October 2024 Progress in Guaranteed Safe AI

Safeguarded AI TA 1.4 funding call, a way of verifying quantitative properties of program pairs, and previews of the recent ICLR submissions.

Oct 28, 2024 • Quinn Dougherty

September 2024

September 2024 Progress in Guaranteed Safe AI

This month we have elicitation of GPT-o1’s Lean4 skills, a worked example in self driving cars, and an interactive proof approach to correctness of…

Sep 26, 2024 • Quinn Dougherty

2

August 2024

August 2024 Progress in Guaranteed Safe AI

$10k on the line, RL from Proof Assistant Feedback, Aria’s TA3

Aug 27, 2024 • Quinn Dougherty

1

July 2024

July 2024 Welcome Edition

A new newsletter about guaranteed safe AI

Jul 25, 2024 • Quinn Dougherty

1

Coming soon

This is Can We Secure AI With Formal Methods?.

Jul 17, 2024 • Quinn Dougherty