Abu Dhabi Startup Synthan Sciences Unveils Safety Architecture for Physical AI

Abu Dhabi-based startup Synthan Sciences has unveiled a proprietary three-layer safety architecture aimed at governing "physical AI" — autonomous vehicles, humanoid robots and other intelligent machines that operate in the real world. Founded by technologist and author George Bancs, the company says its approach fills a "critical gap" in current AI infrastructure by embedding safety at the hardware level, adding communication and behavioral protocols, and providing an identity and certification framework. Synthan Sciences operates under the Abu Dhabi Global Market (ADGM) regulatory framework and is preparing for a seed funding round as it seeks to set foundational standards for the sector. The company points to projections that the global autonomous systems market could exceed $2 trillion by 2030 to underline the economic stakes.

"We're building the seatbelt for the age of physical AI. Safety can't be an afterthought," George Bancs, Founder and CEO of Synthan Sciences, said in a statement distributed via EIN Presswire by BANCS BRANDS LIMITED on April 13, 2026.

Context and technical outline

Synthan Sciences describes its safety stack as a "trust infrastructure" composed of three integrated layers designed specifically for machines that move and act alongside people. The company frames the work at the intersection of AI safety, robotics and regulatory technology, arguing that an infrastructure-first approach is necessary as decision-making shifts from screens into physical environments.

• Hardware-level safety component: A safety element embedded directly into autonomous systems intended to provide fail-safes and baseline protections at the device level.
• Communication and behavioral protocol layer: Standards for how machines communicate intent and coordinate behavior with humans and other systems to reduce ambiguity and unsafe interactions.
• Identity and certification framework: Authentication, provenance and certification mechanisms designed to attribute responsibility and enable auditing of physical AI actors.

"We're entering an era where AI isn't just making decisions on a screen — it's making decisions in the real world, with real consequences," Bancs added, emphasizing that the firm's mission is "to build the safety layer that makes that transition trustworthy." The company positions the safety stack as analogous to cybersecurity for the internet: an infrastructural foundation that other developers, manufacturers and regulators can adopt.

Outlook and next steps

Synthan Sciences is preparing to raise seed capital to advance its technology and to promote adoption of its proposed standards. The startup also ties its theoretical underpinnings to a three-volume book series by Bancs, The Syncyclopedia of Synthanity, which the company says is available on Amazon and through major booksellers. Synthan positions its approach as industry-shaping, seeking to influence how safety, identity and accountability are embedded into the next generation of autonomous systems as markets expand toward the multi-trillion-dollar forecasts cited by the company.