Cardano's news cycle changed character when the network stopped being run by the company that built it. Headlines still describe protocol upgrades as product launches. Under Voltaire they are votes, and the vote record is public.
That shift matters for anyone reading about ADA. A claim about a coming upgrade used to be believed or doubted on the reputation of one engineering firm. Today you can check whether delegated representatives actually approved it, which protocol version it moves the chain to, and whether the activation timestamp has passed. This piece lays out who controls what, which numbers are worth tracking, and the specific places to check before repeating a claim.
Who builds Cardano, and who will after 2026
Charles Hoskinson started the project in 2015, after leaving Ethereum in 2014 over a dispute about whether that network should be structured as a commercial venture. He co-founded an engineering company with Jeremy Wood, originally called IOHK and later renamed Input Output Global. Everyone shortens it to IOG.
The mainnet went live in September 2017. Three organisations were set up alongside each other rather than inside one company. IOG handled core protocol engineering. The Cardano Foundation, a Swiss nonprofit, took stewardship, standards and ecosystem oversight. Emurgo was the commercial arm, tasked with business adoption. Intersect came later, as a member-based body that runs the governance machinery rather than writing the code.
That four-way split is why "Cardano announced X" is almost always a sloppy sentence. Which of the four said it, and do they have the authority to deliver it?
The answer is changing. On 17 July 2026 IOG said it would hand responsibility for core components to independent teams, starting in August 2026 and running through 2027. The list covers the Haskell node, the Plutus smart-contract platform, the Daedalus desktop wallet, Hydra, and developer relations. Two named groups are Se7en Labs, an infrastructure agency with a Solana background, and Teragone, a cryptography team that has led development of Mithril for several years. Three separate client implementations in Haskell, Rust and Go are meant to run in parallel, with Intersect and a body called Pragma approving formal specifications through member votes. IOG said it would refocus on research and new ventures through IO Labs and IO Ventures.
Hoskinson, who is also chief executive of IOG, denied rumours in July 2026 that he was retiring from the project, calling them a complete fabrication spread by edited clips.
Voltaire turned upgrades into votes
Cardano's roadmap was written as five named eras: Byron for the 2017 launch, Shelley for decentralised block production from 2020, Goguen for smart contracts, Basho for scaling, and Voltaire for governance. Voltaire is the one that changed how news gets made, because it moved authority off IOG and onto token holders and their delegates.
The Chang hard fork in September 2024 opened that era. Van Rossem is the clearest example of the new mechanics. It was ratified on 13 July 2026 with 77.63% support from delegated representatives, then activated on 18 July 2026 at 21:44 UTC, moving the chain to Protocol Version 11 and adding Plutus built-in functions that cut the cost of running smart contracts. The project adopted the motto "Built by many, owned by all."
Three things are checkable there, and all three are free. The percentage. The timestamp. The protocol version.
If a post about a Cardano upgrade does not give you at least one of those, it is asking you to take its word for something the network publishes by default. That is the single most useful filter you can apply.
The engineering that has actually shipped
Ouroboros is the consensus protocol, and its main claim to fame is provenance. It was the first proof-of-stake design published with formal, peer-reviewed security proofs, developed with academic cryptographers. ADA holders delegate to stake pools, and pools are chosen to produce blocks in proportion to delegated stake.
Delegation is non-custodial. Your tokens stay in your wallet, there is no slashing, and there is no lock-up period. Rewards have generally sat in the low-to-mid single digits annually, though that moves with network parameters.
The ledger uses an Extended UTXO model, closer to how Bitcoin tracks balances than to Ethereum's account model, except that each unspent output can carry data and a script. Transaction outcomes are deterministic, so you know the result before submitting. Fees are quoted in ADA, are not burned, and flow to stake pool operators and their delegators. Contracts are written in Haskell-based languages, Plutus for general logic and Marlowe for financial agreements.
Scaling runs through several unfinished threads. Hydra creates off-chain "heads" where participants transact quickly before settling to the main chain; benchmarks inside a head have shown high throughput, but production adoption has stayed limited. Mithril uses stake-based signatures to speed up node synchronisation. Leios is the next throughput milestone, with no launch date published. Midnight, a privacy-focused partner chain, reached beta.
Numbers that outlast a headline
ADA has a fixed maximum supply of 45 billion tokens. Its all-time high was about $3.10 in September 2021.
Where it sits now depends on the month you ask. CoinShares put the circulating market capitalisation at roughly $10.5 billion near $0.30 in May 2026. Reporting from July 2026 described ADA around $0.165, a market capitalisation near $6 billion, and total value locked on the network of about $70 million, against more than $4 billion each on Tron and Solana.
That last figure is the one analysts watch, and the one promotional coverage skips. A chain can post perfect uptime and still host very little economic activity.
Cardano's defenders point to a real strength alongside the weak ones: the network has avoided the multi-hour outages that hit several competing layer-1 chains. Conservative, formally verified engineering produces a boring reliability record, and boring is underrated.
Where the criticism is fair
Boring has a cost. The research-first method puts papers before code, which is why Hydra has taken years to reach limited use and why competitors with faster iteration cycles captured most decentralised finance liquidity.
Governance has also been messy in public. Emurgo left the Pentad governance body after a wallet exploit. Investor Justin Bons publicly called for Hoskinson to step down. Hoskinson himself warned earlier in 2026 that deteriorating conditions would push some Cardano-linked projects to shut down, and described the handover as an uncomfortable but necessary stage.
None of that makes the network a scam, and none of it makes it a bargain. It means the interesting question about Cardano in 2026 is institutional: can a community that just inherited the source code keep shipping without the firm that wrote it?
How to verify any Cardano claim yourself
Start with who published it. The IOG blog, the Cardano Foundation site, Intersect at intersectmbo.org, and the community-run Essential Cardano are the primary channels. Secondary outlets are worth reading for context and worth distrusting for specifics.
- Find the entity. IOG, the Foundation, Emurgo and Intersect have different authority, and a claim from one is not a commitment by another.
- Look for the governance record. An upgrade that reached mainnet has a vote and a ratification percentage behind it.
- Check the protocol version. Version 11 arrived with Van Rossem in July 2026, so any current claim should be consistent with that.
- Match the date. Old announcements get recirculated as fresh news constantly.
- Separate the milestone from the price story. A successful hard fork says nothing reliable about what ADA does next.
Language is its own tell. Promises of guaranteed returns, insider access, or a specific percentage move are red flags, as is any post that makes you feel you must act within the hour. Genuine protocol news does not expire.
One more check applies specifically to this chain. Because authority is now split across Intersect, Pragma and outside teams, a claim about "Cardano" doing something should name the body responsible. Vague attribution is a warning, not a detail.
What none of this tells you
Governance quality and token price are different variables. Cardano has shipped a functioning on-chain voting system, a research pipeline with genuine academic output, and an uptime record most competitors would take. It also carries a small decentralised finance footprint relative to its market capitalisation and an unresolved question about whether independent teams can maintain three clients at once.
Both statements can be true at the same time, which is roughly the honest position to hold. Verify the protocol facts against the records named above, treat every price target you read as an opinion, and size any exposure to what you could lose without it mattering.