Rebuff AI
Founded
2023Employees
51-200Funding
AcquiredProducts
0 products listedWebsite
rebuff.ai/
Profile updated
About Rebuff AI
Rebuff is an open-source self-hardening prompt injection detection framework, originally created by Willem Pienaar. The project uses a multi-layered defense combining heuristics, a dedicated LLM, a known-attack database, and canary tokens to protect AI applications.
Rebuff became part of Protect AI in July 2023. Protect AI is a Seattle-based AI security company that was itself acquired by Palo Alto Networks in 2025. Rebuff is maintained as open source under Apache 2.0 alongside other ML security tools.
Specialties
No products listed yet for this company.
Company Timeline
Key milestones in Rebuff AI's history, with sources.
- Acquisition
Palo Alto Networks completes acquisition of Protect AI
Protect AI -- the AI/ML security startup that had owned and maintained the Rebuff open-source project since 2023 -- was folded into Palo Alto Networks' Prisma AIRS platform. Deal reported north of USD500M.
paloaltonetworks.com - Milestone
Rebuff GitHub repository archived (read-only)
protectai/rebuff was archived by its owner; the API record confirms "archived": true and shows no code pushes since August 2024. The project is no longer maintained.
github.com - Acquisition
Protect AI (Rebuff's owner) to be acquired by Palo Alto Networks
Palo Alto Networks announced intent to acquire Seattle AI-security startup Protect AI -- whose prior acquisitions included Rebuff, Laiyer AI, SydeLabs and Huntr -- in a deal reported north of USD500M.
geekwire.com - Acquisition
Protect AI acquires the open-source Rebuff project
Protect AI (maker of the Huntr bug-bounty program) took over the independent Rebuff prompt-injection-detection project and continued its development under the Apache 2.0 license.
darkreading.com - Launch
Rebuff open-sourced as a prompt-injection detection framework
Released as an independent open-source project (not a company) by Willem Pienaar and Shahram Anver, combining heuristics, a dedicated LLM, an attack-pattern vector database and canary tokens to detect prompt injection.
langchain.com
Explore AI Software Reviews
Browse multi-perspective AI panel reviews across hundreds of AI tools, agents, and platforms. Find the right software with insights from all six expert seats — Decision Maker, Domain Strategist, Finance Lead, Domain Practitioner, Power User, and Skeptic.