Mid-2024, we started with building out the Internet of Agents – an infrastructure layer that looks at creating multi-agent systems across vendors, frameworks, orgs and clouds, in an open and interoperable manner. This open infrastructure lets agents discover, collaborate, and transact across vendors. Co-founded AGNTCY — now at the Linux Foundation with 80+ partners including Google, Dell, Oracle and RedHat.
Over the course of working with AGNTCY, A2A, and MCP to plumb these multi-agent systems together, we quickly realized that Connection isn’t Cognition. The Internet of Agents was a great first step – allowing agents to discover and talk to each other. But they are not able to think together, yet. They cannot share intent, coordinate to solve problems or create emergent behaviors.
We are embarking on a journey to create the Internet of Cognition as the platform that allows agents to think together.
Source: SuperDataScience Podcast (Ep. 961) Podcast | Date: Jan 27, 2026
Distributed Artificial Superintelligence
A deep-dive podcast on advancing science with distributed artificial superintelligence, emphasizing the necessity of open multi-agent systems over isolated “AI geniuses.”
Source: IT Visionaries Podcast | Date: 2025/2026
Cisco’s Vijoy Pandey: The New Internet of AI Agents
An interview exploring the foundational communication stacks required to bring order to autonomous agent sprawl.
Source: SuperDataScience Podcast (Ep. 941) Podcast | Date: Nov 18, 2025
Multi-Agent Human Societies
A conversation exploring a future where agents and humans collaborate to make scientific discoveries, alongside an introduction to AGNTCY and the necessity of neutral platforms.
Source: The Pulse of AI | Date: November 2025
The Internet of AI Agents is the Future
A conversation with Jason Stoughton regarding the “Internet of Agents” paradigm, the importance of digital identity, and how startups should navigate the burgeoning agentic ecosystem.
Source: AI in Business Podcast (Emerj) Podcast | Date: Feb 5, 2026
The Internet of Agents and What It Means for Enterprise Leaders
Exploring how enterprises are shifting to probabilistic architectures, highlighting the critical need for Task-Based Access Control (TBAC) to deploy AI safely.
Source: TWIML AI Podcast (Episode 737) | Date: June 2025
Building the Internet of Agents
A technical discussion on the scaffold for multi-agent systems, covering AGNTCY for discovery, identity and access management, observability, syntactic protocols like A2A/MCP and the introduction of SLIM (Secure Low-Latency Interactive Messaging).
Source: Forbes | Date: Mar 10, 2026
The Internet Of Cognition: Building INFRA For Distributed Superintelligence
Today’s AI agents can connect, but they can’t think together. It’s like assembling a team of world-class experts who can only exchange business cards..
Source: Forbes | Date: Feb 6, 2026
The Missing Link In AI Is Getting Agents To Think Together
Scaling artificial intelligence has, until now, largely meant building bigger models But when they try to work together in multi-agent systems, they cannot think together, as human teams can..
Source: VentureBeat | Date: Apr 15, 2026
AI’s next bottleneck isn’t the models — it’s whether agents can think together
When Vijoy Pandey considers the next step for AI agents, his thoughts go back 70,000 years in human history, a critical time in evolution
Source: Axios | Date: Feb 9, 2026
Why AI’s next big leap is collective intelligence
An article discussing the strategic shift from scaling up individual models to horizontal scaling, arguing that isolated agents must learn to “think together” to solve complex enterprise problems.
Source: MIT Technology Review | Date: Feb 6, 2026
Moltbook was Peak AI Theater
Quoted discussing how early attempts at collective intelligence represent our “first attempt at a glider,” establishing the baseline for sustained, distributed superintelligence.
Source: VentureBeat | Date: Jan 29, 2026
AI agents can talk to each other — they just can’t think together yet
Exploring the Internet of Cognition, this piece addresses the practical gap where current protocols allow agents to exchange messages, but fail to provide the semantic layer needed for true coordination.
Source:The Next Platform | Date: Mar 3, 2026
With Cisco Outshift, Agentic AI Is Teed Up For the Internet Of Cognition
Details a vision for the “Internet of Cognition,” exploring how AI must shift from individualistic intelligence to “collective intelligence” by utilizing semantic protocols and a distributed cognition fabric to overcome agent isolation.
Source: arXiv / Cisco Research | Date: Nov 26, 2025
A Layered Protocol Architecture for the Internet of Agents
This academic paper proposes an architecture for the IoA, introducing a novel Layer 8 for communication and Layer 9 for semantic negotiation to establish machine-readable understanding.
Source: Outshift by Cisco | Date: Jan 27, 2026
From connection to cognition: Scaling out superintelligence
This article introduces the Internet of Cognition (IoC), a new architectural approach for distributed artificial superintelligence that enables specialized AI agents and humans to share underlying intent.
Source: Forbes | Date: Feb 25, 2025
Why AI Needs A New Kind Of Internet: The Case For Open Infrastructure
A sponsored feature arguing that our current internet architecture is straining under AI demands, calling for a new, interoperable infrastructure designed specifically for agent-to-agent collaboration.
Source: arXiv / MIT & Cisco | Date: Oct 2025
Ripple Effect Protocol: Coordinating Agent Populations
An academic paper introducing a coordination protocol where agents share sensitivity signals to prevent systemic failures like the “bullwhip effect” in multi-agent supply chains.
Source: Outshift by Cisco | Date: Jan 2026
Scaling Out Superintelligence: Building an Internet of Cognition (Whitepaper)
A foundational whitepaper exploring the shift from connection to cognition, detailing the semantic protocols, cognition fabrics, and engines required for AI agents to negotiate common goals.
Source: Linux Foundation PR | Date: Jul 29, 2025
Cisco donates the AGNTCY project to the Linux Foundation
News coverage detailing the donation of the AGNTCY project to ensure open multi-agent infrastructure remains neutral and accessible to the broader developer community.
Source: Network World | Date: Jan 30, 2025
Cisco touts ‘Internet of Agents’ for secure AI agent collaboration
An overview of the proposed open-sourced, three-layer architecture designed to prevent agent sprawl and enable secure, quantum-safe communication between autonomous systems.
Source: The Next Platform | Date: Jan 28, 2025
The Coming Age Of The Internet Of Agents
An in-depth look at the network infrastructure needed to connect vast numbers of AI agents from multiple vendors, highlighting the shift from deterministic to probabilistic computing.
Source: Outshift by Cisco | Date: Jan 22, 2025
Building a scalable foundation for AI’s future
This blog outlines the foundational vision for the Internet of Agents, describing an open, interoperable layer that allows AI agents to find, verify, and collaborate.
Source: CX Today | Date: Dec 2024
Cisco Outlines Its “Internet of Agents” Vision
Summarizing the vision presented at AWS re:Invent for an interoperable system that prevents isolated “walled gardens” and avoids the dominance of a single monopolistic super-agent.
Source: Outshift by Cisco | Date: 2025
Access control is the biggest AI security risk
Details the imperative of abandoning legacy human-centric IAM in favor of Task, Tool, and Transaction-Based Access Control (TBAC) to secure multi-agent systems.