FIJISHI EPL - ABOUT US
Fijishi was founded in 2023. The thinking behind it started earlier.
In 2007, our founder began an undergraduate education against the backdrop of what would become the 2008 financial crisis - watching institutions that were assumed to be stable fail, and watching the regulatory and compliance frameworks built to prevent exactly that failure prove unable to see it coming or contain it once it arrived.
That observation sat with him for over a decade, through the years that followed the crisis, as the same pattern repeated in new forms - new technologies deployed faster than the governance built to hold them accountable, new frameworks written to describe risk rather than to prevent it. Compliance told institutions what to report. It rarely told autonomous systems what they were permitted to do, or gave anyone the authority to stop them before harm occurred.
By the time artificial intelligence moved from research labs into government infrastructure - welfare systems, financial regulation, defence, courts - that gap had not been closed. It had simply grown larger, and more consequential, because the systems now making autonomous decisions were faster and less visible than the ones that failed in 2008.
Fijishi was built to close it. Not another compliance framework, and not another infrastructure product - a constitutional command layer that specifies, before deployment, what an autonomous system is permitted to do, how a sovereign authority overrides it, and what proof of correct operation looks like when that decision is challenged in a court, a parliamentary committee, or a regulatory proceeding.
The company was formally incorporated in 2023. The doctrine it is built on took longer than that to arrive at.
In 2007, our founder began an undergraduate education against the backdrop of what would become the 2008 financial crisis - watching institutions that were assumed to be stable fail, and watching the regulatory and compliance frameworks built to prevent exactly that failure prove unable to see it coming or contain it once it arrived.
That observation sat with him for over a decade, through the years that followed the crisis, as the same pattern repeated in new forms - new technologies deployed faster than the governance built to hold them accountable, new frameworks written to describe risk rather than to prevent it. Compliance told institutions what to report. It rarely told autonomous systems what they were permitted to do, or gave anyone the authority to stop them before harm occurred.
By the time artificial intelligence moved from research labs into government infrastructure - welfare systems, financial regulation, defence, courts - that gap had not been closed. It had simply grown larger, and more consequential, because the systems now making autonomous decisions were faster and less visible than the ones that failed in 2008.
Fijishi was built to close it. Not another compliance framework, and not another infrastructure product - a constitutional command layer that specifies, before deployment, what an autonomous system is permitted to do, how a sovereign authority overrides it, and what proof of correct operation looks like when that decision is challenged in a court, a parliamentary committee, or a regulatory proceeding.
The company was formally incorporated in 2023. The doctrine it is built on took longer than that to arrive at.