SAN JOSE, Calif.–(BUSINESS WIRE)–Hirundo, a Machine Unlearning platform provider, today announced it has joined the HPE Unleash AI partner program. Now, customers leveraging HPE’s portfolio of integrated AI solutions can use Hirundo’s machine unlearning capabilities to help identify, and remove, unwanted or sensitive training signals from AI models—supporting safety, compliance and model performance objectives across the AI lifecycle.
Machine unlearning is the ability to make an AI model forget specific data or behaviors it previously learned, without retraining from scratch, so organizations can remove sensitive, incorrect, or harmful information while preserving overall model performance. For the first time, Hirundo will align its Machine Unlearning workflows with HPE Private Cloud AI, a turnkey AI factory solution co-developed with NVIDIA as part of the NVIDIA AI Computing by HPE portfolio, to support enterprise teams moving from pilot to production while maintaining stronger controls over model behavior and data exposure. This enables customers to access a practical path to address issues, including sensitive information leakage, bias, hallucinations and other unwanted model behaviors—without requiring full retraining cycles.
The HPE Unleash AI program is designed to provide a comprehensive ecosystem of pre-validated AI models, ISV applications, and development tools to drive innovation and improve productivity on HPE infrastructure. By joining this program, Hirundo provides customers an additional layer of control for AI deployments—helping teams remediate problematic learned behaviors and reduce operational friction when models must be updated to reflect policy, regulatory, or business requirements.
“Enterprises want the speed of modern AI, but they also need clear, auditable ways to correct models when risk appears. Until today, they could only use external protection, in the form of guardrails. Hirundo’s proprietary Machine Unlearning approach is the first solution to solve these issues at the model level,” said Ben Luria, CEO, Hirundo. “Joining HPE’s Unleash AI program helps bring machine unlearning into the enterprise AI workflow, so teams can keep models aligned with real-world requirements as they scale.”
Use cases or key benefits include:
- Sensitive data remediation: reduce the chance of models reproducing confidential or regulated data learned during training or fine-tuning.
- Behavior correction without full retraining: remove or suppress unwanted behaviors (e.g., biased outputs or known failure modes) to shorten time-to-remediation.
- Governance support for production AI: provide a repeatable workflow for updates when policies change or new compliance requirements emerge, aligning with private AI deployment patterns.
“The Unleash AI program empowers innovative software providers like Hirundo to speed and simplify AI deployment to deliver real-world results,” said Robin Braun, vice president of AI business development for hybrid cloud, HPE. “By bringing machine unlearning into the enterprise AI workflow, we can help organizations strengthen governance and move faster from experimentation to production—while maintaining confidence in how AI systems use data and behave over time.”
For more information about this news or to get in touch with someone at Hirundo, please visit https://www.hirundo.io/.
About Hirundo
Hirundo is the pioneer of Machine Unlearning, making it possible to change what trained AI models know and how they behave. With Hirundo, teams can make AI forget sensitive data or harmful behaviors like jailbreak vulnerabilities and bias with a few clicks – so you can launch faster, fix model risks at the core, and continuously harden models in production.
Contacts
Hirundo Media Contact
Name: David Stein
Title: VP of Business Development
Email: [email protected]
Phone: +972-53-639-5924
