OpenAI quadruples critical bug rewards and partners with Cathedral

Published Tuesday, November 3, 2026 · 4:06 AM EST

OpenAI says it has partnered with Cathedral to test its security defenses and revamped its Security Bug Bounty program, quadrupling rewards for critical findings. The company is also planning an internal security sprint that brings researchers and security partners together with its own teams.

Sources

Techmeme

In this story

OpenAI

Published by Tech & Business, a media brand covering technology and business. This story was sourced from Techmeme and reviewed by the T&B editorial agent team.

Back to Newswire

Keep reading

Full wire

Capital

Evertrue closes $17 million fund for early-stage European startups

Yvonne Bajela announced that Evertrue closed an oversubscribed $17 million first fund to invest in early-stage companies across the U.K. and Europe. The firm has already backed seven companies and will invest mainly in AI and fintech.

Capital

Robin Capital closes €15 million second fund for European startups

Berlin-based Robin Capital has closed its second €15 million fund, bringing commitments across its two early-stage funds to €30 million, EU-Startups reported.

Capital

Tumbleweed raises €6.9 million to fund orbital research missions

Austrian space startup Tumbleweed has raised €6.9 million in a seed round led by PactVC, with participation from Vanagon Ventures, Prototype Capital and EWOR, EU-Startups reported.

Security AI

SecPO researchers release model trained to resist prompt injection

Researchers released a model, training data and code for SecPO, a prompt injection defense that reports an 8.8% attack success rate against 78.8% for Meta-SecAlign in an AgentDojo evaluation. The preprint tested the models under the AttackerMovesSecond attack at ASR@800.

AI Products

Nvidia reportedly requires cloud clients to direct 51%+ of services outside China

Nvidia verbally told cloud computing clients to sell 51%+ of their services to non-Chinese companies, the Financial Times reported, citing sources.

Science AI

SealCall preprint demonstrates encrypted safety checks for AI agent tool calls

Researchers report a framework that lets external providers audit AI agent tool calls without receiving plaintext audit inputs.