# With AGI just around the corner, could Tezos use AI to upgrade and improve the protocol on it\`s own?
**URL:** https://forum.tezosagora.org/t/with-agi-just-around-the-corner-could-tezos-use-ai-to-upgrade-and-improve-the-protocol-on-it-s-own/7183
**Category:** Research and Development
**Created:** [September 19, 2026, 4:25pm UTC](https://forum.tezosagora.org/t/with-agi-just-around-the-corner-could-tezos-use-ai-to-upgrade-and-improve-the-protocol-on-it-s-own/7183 "2026-09-19T16:25:54Z")
**Posts on this page:** 4
**Page:** 1
### Author:  [@JASP](https://forum.tezosagora.org/u/JASP)
#### Post date: [September 19, 2026, 4:25pm UTC](https://forum.tezosagora.org/t/with-agi-just-around-the-corner-could-tezos-use-ai-to-upgrade-and-improve-the-protocol-on-it-s-own/7183/1 "2026-09-19T16:25:54Z")
So last 10 years has been a ride for Tezos as anyone that has been around from the the start knows. The protocol itself made a lot of upgrades building a solid foundation but it still lacks a lot.
With all the AI revolution coming up the one feature Tezos protocol has that could be a game changer in my opnion is the possibility of self-amending. This might be the point where the fast and easy way of Tezos to self-amend really comes into play.
Have there ever been any talk on this topic of Tezos Foundation directing some of it`s resources to create a fully AI automated protocol focused on upgrading and improving the protocol based on usage and experience? Tezos protocol could even be a foundation for other projects looking in a way of using AI to build but still keep it under rails and rules (voting system).
Edit: Going with descentralization - the idea would be that the AI is **ownerless and emergent from the protocol itself**. It would not be a product owned by a foundation or Nomadic labs for example - it would be an emergent rule set. Baker approval would be the safeguard of every proposal the AI makes.
---
### Author:  [@weiss](https://forum.tezosagora.org/u/weiss)
#### Post date: [September 19, 2026, 6:39pm UTC](https://forum.tezosagora.org/t/with-agi-just-around-the-corner-could-tezos-use-ai-to-upgrade-and-improve-the-protocol-on-it-s-own/7183/2 "2026-09-19T18:39:22Z")
If we also replace bakers with agents, we can then create Blockchain Skynet. It will then pump itself to the moon and establish the new world order
---
### Author:  [@JASP](https://forum.tezosagora.org/u/JASP)
#### Post date: [September 19, 2026, 9:14pm UTC](https://forum.tezosagora.org/t/with-agi-just-around-the-corner-could-tezos-use-ai-to-upgrade-and-improve-the-protocol-on-it-s-own/7183/3 "2026-09-19T21:14:21Z")
After thounsands of years of corrupt human governance, I dare to be optmistic that given a chance real descentralization may be a game changer. AI might be the path to make that possible actually.
All this pessimism around AI is just the most common human trait expressing itself: Fear.
Anyway - AI will and already is doing better work than developers. It will be in usage to improve many system and softwares in the coming years and I hope Tezos is one of them.
Also - forget about price and moon - we should be mature enough by now to be beyond that. Having a protocol driven partly by AI and voted by human agency may bring some good change.
---
### Author:  [@weiss](https://forum.tezosagora.org/u/weiss)
#### Post date: [September 19, 2026, 11:00pm UTC](https://forum.tezosagora.org/t/with-agi-just-around-the-corner-could-tezos-use-ai-to-upgrade-and-improve-the-protocol-on-it-s-own/7183/4 "2026-09-19T23:00:48Z")
Forget the price and going to the moon? Boooooring, dear sir. If that’s what maturity looks like, count me out. I’m only here to witness the bullish AI!
On a serious note: AI always needs a supervision of competent humans. I’ve seen enough of the cases where latest AI models were planting security issues in the code and then being like “you are totally right!”