If you have seen “IPv8” trending on LinkedIn or Hacker News over the past few months and wondered whether you missed a memo about a new global internet standard, you did not miss anything. There is no IPv8 standard. There is a draft, and the gap between those two words matters a great deal to anyone running production networks.

This piece lays out what IPv8 actually is, what it claims to fix, where the claims hold up, and where they do not. No hype, no dismissal either. Just the facts as they stand today.

What IPv8 Actually is

IPv8 is an Internet-Draft submitted to the Internet Engineering Task Force (IETF) on 14 April 2026 by an individual author, Jamie Thain of One Limited, a Bermuda-registered company. It is titled “Internet Protocol Version 8” and carries the document identifier draft-thain-ipv8-00.

That is an important detail on its own. Anyone can submit an Internet-Draft to the IETF. Submission does not imply review, approval, or endorsement by the IETF. The document itself states plainly that it has no formal standing in the standards process. Internet-Drafts are valid for a maximum of six months and expire automatically if no working group picks them up. This particular draft is due to expire on 16 October 2026 unless it is renewed or adopted.

So before anything else: IPv8 is a proposal from one author, not an approved successor to IPv4 or IPv6, and it may simply lapse.

There is also a naming clash worth flagging. The version number 8 was previously assigned decades ago to an experimental, now-obsolete protocol called PIP. Reusing “IPv8” for an unrelated proposal has added to the confusion around what people are actually discussing when the term comes up.

Why the draft exists

The pitch behind IPv8 is built on a real, well-documented problem: IPv6 has been available for over 25 years and still carries a minority of global internet traffic, while IPv4 addresses ran out at the IANA level back in February 2011. Most operators have coped using Carrier-Grade NAT (CGNAT), which works but adds latency, breaks peer-to-peer connections, and makes troubleshooting harder.

The draft's core argument is that IPv6 solved address exhaustion but never solved the operational pain of running a network, and that the “dual stack” transition model it demands (every device and network segment must speak IPv4 and IPv6 at once, often indefinitely) is what stalled IPv6 adoption for a quarter century. IPv8 tries to route around that failure by promising address expansion and management improvements without a mandatory transition period.

How IPv8 claims to work

Stripped to its essentials, the draft proposes:

A 64-bit address, not 128-bit. An IPv8 address is written as r.r.r.r.n.n.n.n, doubling a familiar IPv4 address rather than replacing it with the longer hexadecimal format IPv6 uses. The first 32 bits (r.r.r.r) encode the Autonomous System Number (ASN) of the network the address belongs to; the last 32 bits (n.n.n.n) are a host address that behaves exactly like an IPv4 address.

A claimed backward-compatibility trick. When the routing prefix is set to all zeros (0.0.0.0.n.n.n.n), the address is defined as being processed under standard IPv4 rules. The draft calls this “IPv4 is a proper subset of IPv8” and uses it to argue that no existing device, application, or network needs to change.

A single “Zone Server” for network management. Instead of running DHCP, DNS, NTP, logging, and authentication as separate systems, IPv8 proposes folding all of them into one platform (DHCP8, DNS8, NTP8, NetLog8, OAuth8, WHOIS8, ACL8, XLATE8) that issues everything a device needs in one exchange, authenticated by OAuth2 JSON Web Tokens.

A route-validation layer. Outbound BGP8 routes and connections would be checked against a WHOIS8 registry before being installed or forwarded, aimed at reducing prefix hijacking, route leaks, and malware command-and-control traffic that relies on hardcoded IP destinations.

A bounded global routing table. By allocating a full 2^32 host addresses per ASN and enforcing a /16 minimum prefix at network boundaries, the draft argues the global routing table would stop growing the way the current BGP4 table has (past 900,000 prefixes with no structural limit).

The claimed benefits, fairly stated

Read generously, IPv8 is arguing for four things:

• More addresses without a forced migration. 2^64 addresses, with every ASN holder getting over four billion, distributed through a mechanism the draft says is “a software update,” not new hardware.
• One coherent management layer instead of a dozen independently configured services with inconsistent authentication.
• Built-in route and traffic validation to reduce spoofing, hijacking, and unauthenticated lateral movement inside a network.
• A structurally bounded routing table, which would matter to carriers and large network operators dealing with BGP table growth.

These are real problems in networking today. Nobody disputes that dual-stack IPv6 migration has been operationally painful, or that DHCP/DNS/authentication fragmentation is a genuine source of misconfiguration and breach risk.

Where it falls apart

This is where the evidence gets uncomfortable for the draft, and it is worth being precise rather than dismissive.

The backward-compatibility claim is contradicted by the rest of the same document. Independent reviewers who read the full draft, not just the abstract, pointed out that it simultaneously requires a new socket API, a new DNS record type, new ARP and ICMP variants, new versions of BGP, OSPF, and IS-IS, mandatory Zone Server infrastructure, mandatory OAuth2 authentication on switch ports, and certified NIC firmware with hardware-enforced rate limits. Any existing IPv4-only router, switch, NIC, or firewall that receives a packet with version number 8 in the header will not parse it correctly, and most will simply drop it. “No modification required” is contradicted on nearly every page of the specification that makes the claim.

There is no vendor, carrier, or standards-body commitment behind it. No router or NIC vendor has announced support. No IETF working group has adopted the draft. It exists purely as an individual submission with an open comment period that, per IETF process, expires if nobody picks it up.

The authorship itself is disputed. The named author, Jamie Thain, is not a recognised figure in IETF or networking-standards circles, and has publicly acknowledged using AI chatbots to help write the draft. Independent AI-detection analysis flagged large sections of the document as very likely AI-generated. That does not automatically make the ideas wrong, but it does mean the draft has not been through the kind of peer scrutiny that normally precedes a serious protocol proposal, and several of its more sweeping claims (such as “100% backward compatible”) do not survive a careful read of the same document's own requirements.

The response from working network engineers has been sharply critical. The proposal generated a large, largely critical discussion thread on Hacker News, and multiple network architects and ISP engineers who examined the document in detail called specific provisions unworkable given how deeply address handling is baked into existing routing, forwarding, and security infrastructure. Even commentary more sympathetic to the underlying goals has stopped well short of endorsing IPv8 as presented.

At the time of writing, IPv8 has no IETF working group, no vendor roadmap, and no operational deployment anywhere. It is a single Internet-Draft with a hard expiry date, being actively crowdfunded by its author to build a demonstration testbed.

What this actually means if you run a network or buy cloud infrastructure

Nothing changes for your infrastructure planning today. That is the honest, slightly less exciting answer.

If you are an enterprise, a regulated financial services entity, or a government-adjacent buyer thinking about IPv6 adoption, data residency, or network security architecture, the calculus has not shifted. IPv6 adoption in India and globally continues to grow steadily for genuine, board-level reasons: address availability, routing efficiency, and increasingly, alignment with government and telecom-driven mandates. Waiting for IPv8 to resolve your addressing or security roadmap would mean waiting on a draft that may not exist as a functioning specification six months from now.

Where the IPv8 conversation is genuinely useful is as a reminder of what actually matters in network and cloud security architecture, regardless of which IP version carries the packet: authenticated device onboarding, consistent identity across every network service, validated egress routing, and a security model that does not depend on any single protocol generation to hold up. Those are design principles, not a version number, and they apply whether your infrastructure runs on IPv4 with CGNAT, dual-stack IPv6, or whatever comes after it.

For organisations evaluating sovereign, India-based private cloud infrastructure under the DPDP Act, 2023, and SEBI's IT governance framework, the more relevant question is not which IP protocol your provider uses, but where your data physically sits, whose legal jurisdiction it falls under, and how tightly authentication and route validation are enforced regardless of protocol version. That conversation does not wait for IETF consensus.

Frequently Asked Questions

• Is IPv8 a real internet standard?
No. It is an individual IETF Internet-Draft submitted in April 2026. It has not been adopted by any IETF working group and has no formal standing. It expires automatically on 16 October 2026 unless renewed or adopted.

• Is IPv8 actually backward compatible with IPv4?
The draft claims full backward compatibility by defining an all-zero routing prefix as a standard IPv4 address. However, the same draft requires new socket APIs, new DNS record types, new routing protocol versions, and certified hardware, all of which existing IPv4 equipment cannot process without upgrades. Independent reviewers consider the “no modification required” claim inconsistent with the rest of the specification.

• Who created IPv8?
Jamie Thain, described as an IT entrepreneur associated with a Bermuda-registered company called One Limited, submitted the draft. He has stated publicly that he used AI tools to help write it and is crowdfunding a testbed to demonstrate the concept.

• Should my organisation plan around IPv8?
Not at this stage. There is no vendor support, no working group adoption, and no deployment anywhere. IPv6 adoption and sound network security architecture remain the relevant priorities today.

• What happened to IPv6 that IPv8 is trying to fix?
IPv6 solved address exhaustion with a 128-bit address space but required dual-stack operation (running IPv4 and IPv6 side by side) for the transition, which proved operationally expensive with no clear forcing function. After 25 years, IPv6 still carries a minority of global internet traffic. IPv8 tries to avoid the dual-stack requirement, though its practical feasibility is disputed.

---

ZeaCloud tracks developments in network protocols, cloud security, and data sovereignty so our customers do not have to separate signal from noise on their own. For a conversation about India-based private cloud infrastructure, DPDP Act compliance, and network security architecture that holds up regardless of which protocol draft is trending this month, reach out to our team.