Artificial intelligence harm reduction company Umanitek AG today announced the launch of Guardian Agent, an AI identity protection platform designed to combat large language model hallucinations, deepfakes, impersonation and threats to digital identities.

It also revealed the formation of a new advisory board and backing from new investors.

Umanitek addresses what it calls a fundamental problem where technology platforms such as TikTok, X and law enforcement agencies cannot share data because of privacy, competition and geopolitical concerns, yet harmful content spreads across all platforms, age groups and devices.

The Umanitek Guardian Agent operates a decentralized infrastructure that allows platforms and law enforcement agencies to signal and verify if content has been flagged as harmful without ever sharing the underlying data, creating what Umanitek calls an effective “immune system for the internet.”

Guardian Agent addresses three critical risk surfaces: hallucinations and misrepresentations from the leading AI models such as ChatGPT, Claude, Grok, Gemini and Perplexity; fake accounts and impersonation across social platforms, starting with TikTok and X; and broad content and IP infringement across those LLMs and platforms.

The company’s platform monitors hundreds of millions of accounts to detect fake profiles using a user’s images, usernames or bios and flags sudden narrative shifts or coordinated attacks in posts and mentions. Users receive live risk scores that quantify their exposure to online threats across these categories in real time.

The Umanitek platform then generates verifiable digital evidence packs that show when and where content appeared, what it contained and how it evolved to reduce investigative and legal costs for users pursuing immediate takedown requests or accelerated legal action.

Guardian Agent offers 24/7 monitoring of what any large or small language model says about an individual or a brand, content or organization. In the case where misrepresentation is detected, evidence material is created to pursue removal, complemented by an automated mechanism that helps to request corrections and hold companies accountable for publishing libel, lies and content infringement.

“We are entering a race toward intimacy where people will talk more to machines than to humans,” said co-founder Chris Rynning. “In this new reality, data isn’t gold anymore — trust is the new gold. When we can no longer distinguish what’s human or real from what’s not, we need a digital immune system for the web.”

Rynning added that “ef we don’t build decentralized infrastructure now, which gives everyone the ability to signal and verify content, gather evidence and request takedowns without forcing people to give up control of their data, we never will. 2026 will be the year we will fully realize that we are living fake lives online without even knowing it.”

To assist with its goal to facilitate a global immune system for the internet, Umanitek has assembled an advisory board that includes investors and thought leaders.

The board includes Bill Tai, early stage tech investor and industry veteran; HRH Prince Khaled bin Alwaleed bin Talal Al Saud, founder and chief executive officer of KBW Ventures LLC; Joe Betts-LaCroix, founder of Retro Biosciences and former Y Combinator partner; Yalda Aoukar, London based fund manager; Jeremy Achin, serial AI entrepreneur and AI fund manager; Mark Love, early Facebook investor and U.S. based multi-asset investor; Victoria Vysotina, UBS Private Wealth Director; Anastasios Economou, Monaco based Family Office and tech investor; Magnus Grimeland, CEO of Antler.vc; and Ismael Sassi, London based fund manager and AI investor.

Along with the launch of Guardian Agent and the advisory board, Umanitk also said that it had raised new funding in a round that included Bill Tai, Magnus Grimeland, KBW Ventures and the investment arm of a Piëch Family Office. The amount of funding raised was not disclosed.

Image: SiliconANGLE/Ideogram