

Copenhagen, Denmark
New security layer screens sensitive data before it reaches AI language models - giving chat users in organizations real-time warnings and application owners full audit trails, and control over what data their AI ever sees
[Copenhagen, Denmark, April 29, 2026]
2021.AI and Safe Online today announced the launch of “Chat Guardian” - a joint data protection offering designed to help organizations use artificial intelligence safely and in full compliance with GDPR regulations.
The new a offering adds a critical security layer to the GRACE AI Platform, automatically screening data for GDPR-sensitive content before it is processed by a large language model (LLM). When sensitive data is detected, users receive an immediate warning and are given the choice to remove, redact, or reroute the information - putting compliance control back in the hands of the user organization.
As AI adoption accelerates across industries, organizations face a growing but often overlooked risk: employees routinely upload documents, paste data, and build AI-powered assistants using information that contains personally identifiable information (PII), confidential client data, financial records, and other GDPR-protected content frequently without realizing it.
Unlike cybersecurity threats that come from outside an organization, this risk is internal and invisible. Standard AI platforms offer no mechanism to detect, flag, or log sensitive data inputs leaving compliance teams with no visibility and organizations exposed to significant regulatory liability.
"Most organizations are focused on what AI outputs. We think the bigger risk is what goes in," said [Bjørn Olesen, Sales Director, 2021.AI]. "Chat Guardian closes a compliance gap that every AI-adopting organization needs to address — and it does so without slowing teams down."
While Chat Guardian is designed for any organization using AI with real business data, it is particularly relevant for industries operating under strict confidentiality and data protection obligations - including companies in the critical infrastructure and the public sector.
For organizations in regulated industries, the stakes are especially high. Sharing sensitive documents, client records, or proprietary data with an AI tool without adequate screening creates risks that extend beyond GDPR — touching regulatory compliance, professional conduct obligations, and the trust of the clients and stakeholders they serve.
Yet the answer is not to restrict AI - it is to enable it responsibly. The true value of AI comes from assistants that people actually adopt and use freely in their daily work. The added security layer acts as a guardian, not a gatekeeper - giving organizations in regulated industries the confidence to let their users work freely with AI, knowing the right safeguards are already in place.
"Companies today are using AI like never before, and the vast majority take their responsibility for personal data seriously — they genuinely want to get it right. The problem is that, until now, they've had no way to see or control what data their employees are actually sharing with the AI models. With GRACE and our Chat Guardian, organizations finally get the visibility and control they need to embrace AI — without compromising GDPR or the trust of their clients," Christian Bohn, CEO, Safe Online
2021.AI is a European AI governance platform helping organisations scale AI ambitiously without compromising on security and responsibility. Through the GRACE AI Platform, 2021.AI brings governance, risk management, and operations together in one place - giving public institutions and private companies full ownership of their data and AI, and the tools to deploy it in a robust, compliant, and responsible way.
For more information, visit www.2021.ai
Safe Online is a Danish SaaS company focused on automated privacy and data management for businesses. We help organisations uncover and manage personal data across emails, cloud platforms, and now local systems through solutions such as DataMapper. By replacing manual processes with automation, we enable companies to continuously reduce data risk, lower operational workload, and maintain control of their data without the need for significant internal resources.
For more information, visit www.bysafeonline.com
2021.AI
Bjørn Olesen
bjo@2021.ai
Safe OnlineMark Søjnæsmso@safeonline.dk
Stay up to date on our latest news and industry trends