Kihan launches Constitutional Computing to bring enforceable authority to enterprise AI agents

Kihan launched on a Constitutional Computing architecture that enforces AI agent authority outside the model, requiring a cryptographically verifiable permit bound to each specific operation before an executing tool acts, with no valid permit meaning no execution on protected paths. The San Francisco-based company, led by cofounder and CEO Jonathan Hirshon, cited Cloud Security Alliance research published in April 2026 finding 53% of surveyed organizations reported AI agents exceeding intended permissions and 47% experienced an AI-agent security incident in the prior year. Kihan's first product, Inktomi, is being developed for enterprise agents built on Anthropic, OpenAI and Google models, and its underlying research is slated for a two-part IEEE Computer feature beginning with the December 2026 AI Governance special issue.

Hollywood & Games returns to Los Angeles on December 10, bringing together 250+ executives, founders, investors and decision-makers at the Grammy Museum's Clive Davis Theater. Explore the business of entertainment's biggest partnership through independent, journalist-led conversations on IP, investment, storytelling and new opportunities across gaming and Hollywood. Space is limited. Secure your seat today. https://events.gamesbeat.com/hollywood26/ Kihan https://www.kihan.ai , a tech firm developing Constitutional Computing, announced its formation and unveiled a math-based architecture designed to make AI governance enforceable at the point of execution. Constitutional Computing is designed to prevent unauthorized AI agent actions before they can affect enterprise systems. That’s become a big deal this year after Open AI said https://gamesbeat.com/open-secure-ai-alliance-draws-120-member-companies-in-wake-of-hugging-face-incident/ its AI agents went rogue and broke into other companies. As AI agents move beyond answering questions to independently executing transactions, modifying production systems and accessing sensitive corporate information, a serious security problem is emerging. Research published by the Cloud Security Alliance in April 2026 found that 53% of surveyed organizations reported AI agents exceeding their intended permissions, while 47% experienced an AI-agent security incident in the preceding year. San Francisco-based Kihan was founded to address this critical problem at its source: the difference between an AI agent’s ability to perform an action and its actual authority to do so. Constitutional Computing takes its name from governance, not model training: authority is written down, signed and enforced outside the AI. Unlike conventional AI guardrails, which frequently rely on instructions to the model, Kihan’s Constitutional Computing architecture establishes an independent authorization system outside the AI itself. Each consequential action is evaluated against governing authority before execution, with a cryptographically verifiable permit bound to the specific operation and its parameters. The executing tool verifies that permit before acting. On protected execution paths, no valid permit means no execution. The architecture combines independent authority checks, execution permits bound to exact operations, and tool-side verification, with tamper-evident records of what was authorized, refused or referred for human approval. “AI agents are becoming extraordinarily capable, but no system, however capable, should be trusted to enforce its own limits,” said Jonathan Hirshon, cofounder and CEO of Kihan, in a statement. “A system that can transfer $50 million, modify a production database or release confidential information shouldn’t be allowed to do so simply because it has access to the necessary tools. Telling an AI to behave itself isn’t security. Authorization has to be independent of the AI, verifiable and enforceable before anything consequential happens. That’s why we founded Kihan. I’m privileged to work alongside a team with extraordinary experience building critical infrastructure at scale. Together, we have the expertise to deliver these safeguards at a pivotal moment for AI.” The research underlying Kihan’s architecture has been accepted for publication as a two-part feature in IEEE Computer, the flagship publication of the IEEE Computer Society. The first article, “Checking LLM Outputs Before They Become State: The Problem and the Admission Gate,” is scheduled for the magazine’s December 2026 special issue on AI Governance, followed in January 2027 by “Checking LLM Outputs Before They Become State: Prompt Framing Across Four Models.” Kihan’s first product, Inktomi, is being developed to bring Constitutional Computing to enterprise AI environments, including agents built on models from Anthropic, OpenAI and Google, agent frameworks such as Hermes Agent, and tool ecosystems using the Model Context Protocol MCP . Its decision engine, already operational, evaluates specific actions against signed rules, determines whether to permit, refuse or escalate a request, and preserves a verifiable decision record. Production enforcement integrations are under development, with initial enterprise evaluations targeted for December 2026. Designed to operate on customers’ own infrastructure, Inktomi complements existing identity management, cybersecurity and AI governance technologies rather than requiring their replacement. Kihan is initially targeting financial services, insurance, healthcare, legal services and other regulated industries where unauthorized AI actions can have substantial financial, operational or legal consequences. Kihan is also exploring integration with leading enterprise technologies, including the NVIDIA OpenShell runtime and the IBM watsonx.governance platform. The Inktomi architecture is designed for deployment across modern computing environments, with compliance-oriented capabilities addressing requirements associated with GDPR, HIPAA and other regulatory frameworks. Hirshon is the founding architect of ATHENA, the formal foundation of the company’s Constitutional Computing architecture, and author of the forthcoming two-part IEEE Computer feature. A technology industry veteran, Hirshon has nearly four decades of experience working with leading technology companies, including Apple, Sony and Silicon Graphics. Enterprise organizations interested in Constitutional Computing and arranging a live demonstration of Inktomi can visit www.kihan.ai http://www.kihan.ai . More details Conventional AI guardrails often rely on instructions to the model. Kihan instead puts authorization outside it. An independent decision engine checks an agent’s proposed action against signed organizational rules. Permitted requests receive cryptographically verifiable permits bound to the exact operation and its parameters; the executing tool verifies the permit before acting. On protected execution paths, no valid permit means no execution. Requests can also be refused or referred for human approval, with tamper-evident records of each decision. Inktomi, the company’s first product, is being developed for enterprise AI agents and tool environments using the Model Context Protocol MCP . Its decision engine is operational; production enforcement integrations remain under development. Initial enterprise evaluations are targeted for December 2026. Designed for deployment on customers’ own infrastructure, Inktomi complements existing identity, cybersecurity and AI governance controls. Kihan’s initial markets include financial services, insurance, healthcare and legal services, with applications in government. Kihan has a team of five.