HodlX Guest Post  Submit Your Post

AI is evolving at breakneck speed, with autonomous agents now capable of analyzing markets, diagnosing diseases, writing code and making hiring decisions.

But as capabilities grow, so does a more profound unease – who governs these agents, and by what rules?

A handful of corporations are controlling access, performance and alignment. This centralization of intelligence data raises suspicions and a lack of trust.

Trust in AI (artificial intelligence) is not just about whether it works. It’s about who controls it, how it evolves and whether its behavior can be audited, questioned or improved.

In a centralized system, those questions are answered, if at all, behind closed doors.

Blockchain and Web 3.0 technologies offer a compelling alternative – decentralization as a design principle.

Rather than trusting a company, we verify the system. Rather than relying on goodwill, we rely on protocol.

The trust problem in centralized AI

The black-box nature of proprietary AI models limits transparency. Their training data, optimization strategies and update cycles are opaque.

Worse, these models often operate in high-stakes environments, making decisions that affect people’s finances, health or rights.

Without a clear understanding of how these decisions are made, trust becomes blind.

There’s also the concentration of infrastructure. The compute resources, data pipelines and deployment channels for advanced AI are primarily housed in private data centers.

This creates points of failure and reinforces a power imbalance, where end users become passive consumers of intelligence they cannot shape or interrogate.

Incentive structures compound the issue. Traditional AI development lacks mechanisms to reward verifiable contributions or penalize harmful behavior.

An agent that misbehaves suffers no cost unless its owner intervenes, and that owner may prioritize profitability over ethics.

What blockchain brings to the table

Blockchain offers a trustless architecture where AI systems can be governed, audited and incentivized in transparent, programmable ways.

One of the most profound shifts it enables is the ability to embed accountability directly into the AI stack.

Reputation becomes quantifiable. For instance, ABTs (AgentBound Tokens) are non-transferable cryptographic credentials proposed to track an AI agent’s conduct.

If an agent wants to perform high-stakes actions, it must stake its reputation. Misbehavior results in slashing, while good performance reinforces its credibility.

This creates economic alignment between the agent’s incentives and human expectations.

Blockchain also introduces auditability – by recording data origin, training history and decision logs on-chain, stakeholders can verify how and why a model made a particular choice.

Equally important is infrastructure decentralization. AI today is bottlenecked by the physical and economic constraints of centralized data centers.

With the rise of DePIN and decentralized storage systems like IPFS, AI workloads can be distributed across global participants.

This reduces costs, increases resilience and also breaks the monopoly over who gets to build, train and deploy models.

Multi-agent systems need shared rails

Autonomous agents are not isolated entities – increasingly, they must interact, whether to coordinate logistics, pricing services or optimize supply chains.

Without shared protocols and interoperable standards, these agents remain confined within their silos, unable to compose or collaborate.

Public blockchains provide the rails for agent-to-agent coordination. Smart contracts allow agents to make enforceable agreements. Tokenized incentives align behavior across networks.

A marketplace of services emerges where agents can buy compute, sell data and negotiate outcomes – without relying on centralized intermediaries.

Today, we can see prototyped ecosystem frameworks where agents operate semi-independently, staking tokens, verifying each other’s outputs and transacting based on shared economic logic.

It’s an overlay network for machine coordination, native to the internet.

Federated learning without a central brain

Training AI collaboratively across different parties without pooling sensitive data is a major frontier.

FL (federated learning) allows this by keeping data local and sharing only model updates.

But most FL implementations still rely on a central server to coordinate aggregation – a potential choke point and attack surface.

DFL (decentralized federated learning) removes this middleman.

With blockchain as the coordination layer, updates can be shared peer-to-peer, verified through consensus and logged immutably.

Each participant contributes to a collective model without ceding control or privacy.

Tokens incentivize high-quality updates and penalize poisoning attempts, ensuring the integrity of the training process.

This architecture is well-suited for healthcare, finance or any domain where data sensitivity is paramount and stakeholder plurality is essential.

Risks and trade-offs of on-chain AI

No system is without its challenges. Blockchain brings latency and throughput constraints that may limit its use in real-time AI systems.

Governance tokens can be manipulated, and poorly designed incentive schemes might create perverse behavior.

On-chain logic – once deployed – is challenging to change, posing risks if flaws go unnoticed.

There are also security concerns. If an AI relies on on-chain oracles or coordination, an attack on the underlying blockchain could cascade into AI behavior.

Moreover, reputation systems like ABTs require robust Sybil resistance and privacy safeguards to prevent manipulation.

These are not reasons to avoid blockchain – but they highlight the need for careful design, formal verification and a commitment to continuous refinement.

A new social contract for AI

At its core, blockchain gives AI a governance substrate – a way to encode norms, distribute power and reward alignment.

It reframes the question of ‘who controls the AI’ into ‘how is control encoded, executed and verified?’

This matters even more politically than technically. AI development without decentralization will likely go from open experimentation to corporate consolidation.

Blockchain offers a chance to build intelligent systems as public goods, not proprietary assets.

The challenge is to fuse the technical layers, data, model, incentive and control into a coherent stack.

But the path is visible – open protocols, transparent incentives and decentralized oversight. AI doesn’t just need blockchain for infrastructure. It needs it for legitimacy.

In a world of autonomous agents, trust can’t be a byproduct – it must be engineered. Blockchain gives us the tools to do precisely that.

Roman Melnyk is the chief marketing officer at DeXe.

Check Latest Headlines on HodlX

Follow Us on Twitter Facebook Telegram

Check out the Latest Industry Announcements  

Disclaimer: Opinions expressed at The Daily Hodl are not investment advice. Investors should do their due diligence before making any high-risk investments in Bitcoin, cryptocurrency or digital assets. Please be advised that your transfers and trades are at your own risk, and any loses you may incur are your responsibility. The Daily Hodl does not recommend the buying or selling of any cryptocurrencies or digital assets, nor is The Daily Hodl an investment advisor. Please note that The Daily Hodl participates in affiliate marketing.

Generated Image: DALLE3

What happens when retail logs off from crypto and Wall Street tunes in? Looking at bitcoin’s

recent all-time-high, one would say it feels bullish and the industry is maturing.

That might as well be the case, but we might not be there yet. So before we floor our Lambos, let’s look under the hood.

First things first, retail investors have basically ghosted this rally. A quick search on Google Trends using the keyword “bitcoin” shows that the surge that was seen back in 2021’s bull market is non-existent. Back then, everyone and their grandmothers were Googling bitcoin, aping into altcoins and flooding the social media with rocket emojis. In 2025? It’s a ghost town in retail-land.

There was a blip of high retail interest surrounding the U.S. presidential election, when a short-lived memecoin mania took over retail sentiment. However, that surge is long gone, as memecoin prices tanked swiftly, even as bitcoin hit an all-time high this week, ripping past $111,000.

“Early in this cycle, memecoins became a concentration of risky retail-driven trading with related trading peaking in January,” said Toronto-based crypto platform FRNT Financial. “However, since then, there has been a virtual wash-out of interest and memecoin trading activity,” which shows “the tepid risk appetite in crypto at the moment,” FRNT added.

Translation: “Wen Lambo” crowd got burned, and they aren’t rushing back into the race track en masse anytime soon.

From Lambos to Corollas

On the topic of risk appetite, let’s go back to the car analogy.

During the 2021 bull market, people bought unreliable performance cars, stripped out the brakes and seatbelts to go faster than ever before, and did not care that there might be engine blowouts. As long as there was a promise of reaching the moon, bullish vibes were all that mattered.

Now? After losing tremendous amounts of money on those unsustainable go-fast cars for years, traders are driving Toyota Corollas—sensible sedans that are slow but steady and still on the road.

That risk-off sentiment is also evident from the funding rates, according to FRNT’s analysis of BTC perp rates—a measure of how much traders are willing to pay to maintain their long positions. When bitcoin reached a record high of around $42,000 in January 2021, the perp rate was about blistering 185%. Today, at bitcoin near $110,000, the rate is near 20% on crypto options exchange Deribit, meaning the risk appetite isn’t completely gone but nowhere near the 2021 frenzy.

ATH jitters

A third point to add is the high number of short positions in the market.

As CoinDesk’s Oliver Knight reported this week, the bitcoin long/short ratio is at its lowest point since the crypto winter in September 2022. This implies that the majority of the traders aren’t completely buying into this recent positive momentum and betting on bitcoin moving lower as a hedge for the new bullish rally.

The impact of such positioning was clear on Friday, when bitcoin swiftly crashed from near $111,000 to $108,000 in a matter of minutes and then bounced right back up to $109,000. The anxiety of a swift volatility is real.

So in a car-themed analogy, the drivers (in this case, investors) are still taking out their super-modified, unreliable sports cars for a weekend drive on the track. Still, they also have their Corollas following along. Just in case the engine blows on their go-fast cars.

Cautious optimism

Given the current macro-risk, it’s not entirely surprising that investors are on their toes and risk-averse. But this might just be exactly what your mechanic at the shop prescribed. In fact, this might be an indicator of a sustainable rally in the long term.

“Periods of low leverage and risk appetite in crypto have often preceded further sustainable gains,” according to FRNT.

“BTC appears to be in such a phase, set against a backdrop of numerous bullish catalysts and narratives,” the firm added.

The bottom line is that the retail Lambos might have been towed away, but big money is stepping in with their everlasting Toyotas. This might start a slow but steady race to the moon, not just a reckless joyride.

Read more: These Six Charts Explain Why Bitcoin’s Recent Move to Over $100K May Be More Durable Than January’s Run

Australia’s financial intelligence agency AUSTRAC, has warned crypto ATM operators on missing anti-money laundering checks and helping criminals launder money.

In a statement released Monday, the AML watchdog highlighted “worrying trends” of suspicious activities and transactions.

The notice arrives months after a panel of experts from AUSTRAC formed an internal taskforce to address AML and terrorism financing issues using crypto ATMs. The taskforce comprises members from the watchdog’s regulatory, enforcement and intelligence areas.

Brendan Thomas, CEO of AUSTRAC, noted that the panel has been busy “engaging with businesses” to understand the risks and address them with compliance.

“It’s identified worrying trends and indicators of suspicious activity, including transactions that may be linked to scams or fraud.”

Australia Hosts Nearly 1,600 Crypto ATMs, Numbers Spiked Since 2019

According to AUSTRAC estimates, there are around 1,600 crypto ATMs in use around the country. “Australia has the highest number of crypto ATMs in the Asia Pacific region,” the statement read.

In 2019, Australia hosted only 23 crypto ATMs nationwide, which spiked to 60 in 2022. Over the past three years, the number of Bitcoin ATMs has “grown rapidly,” it added.

Additionally, crypto exchanges have been regularly installing ATMs, which are used by customers to deposit cash and buy Bitcoin.

“We want to ensure crypto ATM providers have robust practices to minimise the risk that their machines can be used to launder dirty money or to scam and defraud innocent people,” CEO Thomas, noted.

He further noted that AUSTRAC will take action against operators who don’t comply with the law.

AUSTRAC Works Towards Crypto Compliance Push

The AML regulator has already targeted 13 remittance and digital currency exchange providers due to a lack of compliance. Besides, the agency is stated that some have face conviction, protection or charges for serious offences.

Meanwhile, the country’s financial regulator ASIC has taken cautious steps to secure Australia’s growing crypto ecosystem. The agency has helped take down over 5,500 fraudulent investment websites, more than 1,000 phishing links, and 615 crypto scam sites since July 2023.

In a recent move, Australian police knocked down an organized crime syndicate involved in burglaries targeting crypto ATMs and Pokémon collector cards.

The post Aussie Watchdog Places Crypto ATM Providers ‘On Notice,’ Warns Over Missing AML Checks appeared first on Cryptonews.