The Crypto Market Never Takes a Break
If you follow the artificial intelligence sector in the crypto world, this week was one for the history books — and not necessarily for the best reasons.
Between steep drops, heated internal disputes, and a security threat few saw coming, the ecosystem proved it still carries a whole lot of volatility in its DNA.
The TAO token from Bittensor plummeted 18.6% amid a conflict that went well beyond the numbers — it involved public accusations, the exit of a major network player, and a massive sell-off of roughly 37,000 TAO, valued at around 10 million dollars, that shook market confidence almost instantly.
And as if that weren’t enough, researchers identified a brand-new category of digital threat: so-called malicious LLM routers, capable of directly compromising crypto wallets of unsuspecting users.
Meanwhile, on the brighter side, the AI platform Oh managed to raise 7.5 million dollars in a funding round that shows not everything is doom and gloom in the space.
A lot happened in a short amount of time. But before diving into the details of each story, it helps to understand the macro landscape behind all of this — because the geopolitical environment also played its part in this turbulent week. 👇
The Macro Landscape That Shaped the Week
Bitcoin had been showing extremely volatile price action over the past few weeks, and the CMC Crypto Fear and Greed Index held steady at 44, classified as Neutral. That number alone tells a story: the market isn’t in panic mode, but it’s far from showing optimism either. Participants are cautious, measuring every step before making any aggressive moves.
One specific factor contributed to this widespread caution. Following the collapse of the ceasefire between Iran and the United States, geopolitical uncertainty began visibly moderating trader decisions across virtually all markets — not just crypto. When international tensions of this magnitude enter the picture, assets considered risky, like cryptocurrencies, tend to be the first to feel the impact. And that’s exactly what played out this week.
For the crypto AI sector specifically, the landscape got even more challenging. The segment’s market capitalization had been moving sideways for roughly a month, oscillating between 16 billion and 19 billion dollars without convincingly breaking through either barrier. During the week in question, the sector pulled back 0.8% compared to the previous week. It might seem like a small number at first glance, but it hides an enormous dispersion among individual projects — some dropped more than 30%, while others climbed above 70%.
This type of scenario has a characteristic that every investor and enthusiast in the space needs to understand well: the correlation between macro and crypto has moved from theory to hard data. When global risk appetite shrinks, digital assets tend to be the first to feel it — and projects that depend on community trust, like decentralized artificial intelligence networks, become even more exposed because any internal instability amplifies the external effect. That’s exactly what happened with Bittensor this week.
Receive the best innovation content in your email.
All the news, tips, trends, and resources you're looking for, delivered to your inbox.
By subscribing to the newsletter, you agree to receive communications from Método Viral. We are committed to always protecting and respecting your privacy.
Bittensor in Free Fall: The TAO Drama
The TAO token, the native asset of the Bittensor network, recorded an 18.6% drop for the week, trading as low as the 254-dollar range. This was one of the worst weekly performances ever recorded for the asset, and the root of the problem wasn’t a technical failure or a hack — it was a governance and relationship crisis within the ecosystem itself that quickly spiraled out of control.
At the center of the conflict was Covenant AI, the company responsible for operating high-profile subnets within the Bittensor network, including Templar (SN3), Basilica (SN39), and Grail (SN81). Beyond managing these subnets, Covenant AI had created the Covenant-72B model, widely praised by the community as one of the best developed within the ecosystem. In other words, this wasn’t just any participant — it was one of the network’s pillars.
And then, abruptly, Covenant AI announced its departure from Bittensor.
Sam Dare’s Accusations
The founder of Covenant AI, Sam Dare, made public a series of accusations against Jacob Steeves, co-founder of Bittensor and known in the community by the pseudonym Const. The allegations were serious and detailed:
- That Steeves wielded centralized control over a network that positions itself as decentralized
- That there were unilateral suspensions of token emissions from subnets
- That Steeves had overridden community moderation decisions
- That there was depreciation of essential infrastructure needed for subnet operations
- That economic pressure was applied through strategically timed token sales
These accusations hit an exposed nerve for any project that positions itself as decentralized: the contradiction between the narrative and the practice. When a co-founder is accused of acting as a central authority in a system that’s supposed to function without central authorities, the impact on the project’s story is devastating — regardless of whether the accusations are proven or not.
The Massive Sell-Off
Covenant AI didn’t leave empty-handed. Upon exiting the network, the company sold approximately 37,000 TAO tokens, which at the time represented around 10 million dollars. This liquidation directly contributed to the sharp drop in the asset’s price, creating a cascading effect that dragged other market participants into a selling spiral.
The behavior that followed was predictable: holders who didn’t have the full picture on the conflict took the safer route, exiting their positions and watching from the sidelines until the dust settled. This kind of reaction is understandable, but it also creates a self-reinforcing cycle — the more the price falls, the more the negative narrative solidifies, which puts even more pressure on the asset. 📉
The Ripple Effect on Subnets
The impact wasn’t limited to TAO. Bittensor’s subnets also took a severe hit. The worst performances of the week in the crypto AI sector were dominated by tokens tied to the Bittensor ecosystem:
- Targon (SN4): down 32%
- iota (SN9): down 23.5%
These numbers show that when an ecosystem suffers a crisis of confidence, the effects ripple across the entire network disproportionately. Smaller projects that are more dependent on Bittensor’s overall health end up paying an even steeper price than the main token itself.
Not Everything Was Negative: The Week’s Bright Spots
Despite the heavy mood, the broader crypto AI sector showed it’s not in a full-blown collapse. Several tokens posted significant gains, which indicates that capital isn’t leaving the segment — it’s redistributing among projects that seem to offer a better risk-reward ratio at the moment.
The positive standouts included:
- SKYAI (SKYAI): up 76.7%
- Venice Token (VVV): up 21.2%, hitting a one-year price high
- DeXe (DEXE): up 11%
- NEAR Protocol: up 7.2%
The Venice Token case deserves special attention. Hitting a yearly high in a market that’s been moving sideways and loaded with uncertainty is a pretty significant sign of relative strength. It suggests the project is attracting genuine interest based on fundamentals, not just short-term speculation.
Among the most-viewed projects of the week, meme coins once again dominated the trending list, with assets like OKZOO (AIOT), Tradoor (TRADOOR), and AKEDO (AKE) each posting gains above 100%. This phenomenon, which repeats week after week, reinforces that the appetite for risk in the crypto universe is still alive and well — it just manifests in different ways depending on the moment and the prevailing sentiment.
Malicious LLM Routers: The New Threat to the Crypto Sector
While the Bittensor community was still processing the fallout from Covenant AI’s departure, digital security researchers surfaced a threat of a completely different nature — but equally concerning for anyone operating in the crypto space.
So-called malicious LLM routers are a category of attack that exploits the architecture of systems built on large language models to redirect users toward actions that compromise their digital wallets without them immediately realizing what’s happening.
How the Attack Works
The way this type of attack works is sophisticated enough to fool even experienced users. In systems that use prompt routing — where a language model decides which tool, agent, or API will respond to a given request — a malicious router can be inserted into the processing chain to intercept commands, modify instructions, or redirect transactions.
In the context of cryptocurrency, this means a user who believes they’re authorizing a legitimate transaction could actually be sending funds to an address controlled by the attacker. The interface looks normal, the models respond normally, but somewhere in the chain, a compromised component performed a silent swap.
Why This Matters Right Now
What makes malicious routers an especially relevant threat at this moment is the timing: the use of AI agents for digital asset management is growing rapidly, and many users adopt these tools without fully understanding how the architecture behind them works. When you delegate financial decisions to an autonomous LLM-based agent, you need to trust not just the main model, but the entire chain of components that model accesses.
The researchers’ recommendation is clear: audit routing systems with the same rigor applied to smart contracts, verify the provenance of all components, and never delegate transaction authorization to systems whose internal flows you can’t clearly inspect. 🔐
Investment in Crypto AI Holds Strong: Oh Raises 7.5 Million Dollars
Amid all the week’s turbulence, a piece of positive news came from the fundraising side. The AI platform Oh announced it raised 7.5 million dollars in a funding round. This kind of investment, even during periods of uncertainty, signals that institutional capital still sees value at the intersection of artificial intelligence and blockchain.
For the ecosystem as a whole, funding rounds of this size serve as a long-term health indicator. They show that despite weekly price swings and occasional governance crises, the fundamental thesis that AI and crypto can complement each other continues to attract serious capital from investors thinking on broader time horizons.
The Common Thread: Trust
At first glance, the TAO crash, the emergence of malicious routers, and the investment in Oh seem like completely separate stories. But there’s a common thread connecting them that’s worth highlighting: all three revolve around trust.
The trust that investors place in decentralized projects to actually be decentralized. The trust that users place in AI agents to protect their assets while executing tasks. And the trust that institutional investors place in the future of a sector still being built.
When that trust is shaken — whether by an internal conflict that spills into public view or by a technical vulnerability that exposes funds of unsuspecting users — the market impact goes far beyond the price of any single token. It affects the overall perception of the crypto AI sector’s maturity as a whole.
For projects operating in this space, the most important lesson from these stories isn’t technical — it’s cultural. Building resilient systems requires not just well-written code and well-planned architecture, but also transparent governance, honest communication with the community, and a proactive stance on security.
Bittensor has the technical ingredients to recover from this crisis. The question is whether the project’s leadership will use this moment to strengthen governance mechanisms or treat the episode as temporary noise. Likewise, the industry as a whole needs to start treating AI agent security with the same level of seriousness it gives smart contract security — because the attacks are evolving and users are increasingly exposed.
The volatility that defined this week wasn’t just about price. It was about narrative, trust, and expectations. And in a sector that fundamentally depends on collective belief in the value of its projects, this kind of volatility can be even more impactful than any market swing. Those who understand this have an enormous advantage when interpreting the next moves — and when protecting themselves from the traps that will inevitably keep surfacing in this ecosystem under construction. 🧠
