The latest gains came after Trump mulled sending troops to Kharg Island and urged NATO to help with the war against Iran.

Bitcoin’s price is on the move on Monday morning, surging to a six-week peak of just over $74,000. This might be rather unexpected as the weekend was quite eventful on the Middle East war front, as the US hit a key Iranian island, and legacy financial markets opened hours ago in reaction to the news.

Many altcoins have produced even more impressive increases, including ETH, which has finally climbed above $2,200.

Crypto Market Moves Higher

The total crypto market cap has added over $80 billion to $2.6 trillion on CG as of now. Bitcoin exceeded $74,000 minutes ago, where it faced some resistance, and now sits just below that level. The asset fell toward $70,000 over the weekend after Trump announced “the most powerful bombing raids in Middle East history,” when the US military attacked Kharg Island.

However, it bounced off and eyed $72,000 yesterday, but once the legacy financial markets started to open on Sunday evening and Monday morning, it jumped to the aforementioned six-week peak. ETH is among the top performers in the past 24 hours, surging by 8% to nearly $2,300 for the first time since early February.

Notable gains are evident from ADA (10%), DOT (12%), PEPE (15%), ETC (9%), and others. The total value of wrecked positions has risen to $350 million, according to CoinGlass, with nearly $300 million coming from shorts. Interestingly, ETH shorts are responsible for the largest portion, followed closely by BTC’s.

Latest Developments

Although Trump said at first that the US doesn’t want to attack any of Kharg Island’s oil infrastructure, he threatened to do so if Iran interferes in any form with the “free and safe passage of ships through the Strait of Hormuz.” He later urged numerous countries to send warships to defend the passage.

More recently, he mulled the idea of sending troops on the ground to seize the key island, which is responsible for over 90% of Iran’s oil production.

You may also like:

BREAKING: President Trump is considering putting boots on the ground to seize Iran’s Kharg Island, per Axios.

Details include:

1. The move appears to be contingent on if tankers remain bottled up in the Persian Gulf

2. Trump is working to assemble a coalition of countries to…

— The Kobeissi Letter (@KobeissiLetter) March 16, 2026

Advertisement

The other big development in the past few hours came when Trump suggested that the US has continuously helped NATO with the war in Ukraine, so the alliance should return the favor now, at least with the Strait.

Artificial intelligence has quickly become one of the hottest narratives in crypto trading. From automated trading bots to fully autonomous AI agents scanning blockchain data in real time, many believe AI could unlock the holy grail of trading: consistent market prediction.

But can AI truly predict crypto markets better than traditional strategies? Or is much of the excitement driven by hype rather than proven results?

Let’s break down the reality behind AI-driven crypto trading.

The Rise of Machine Learning Models in Crypto

Machine learning models are increasingly being used by traders, hedge funds, and algorithmic platforms to analyze massive amounts of market data. Unlike traditional trading systems that rely on fixed rules, machine learning models continuously learn from historical patterns and adapt to new information.

Some of the most common AI models used in crypto trading include:

1. Time Series Forecasting Models

These models attempt to predict future prices using historical market data such as:

• Price movements

• Trading volume

• Order book depth

• Volatility patterns

Techniques like LSTM neural networks, ARIMA models, and transformers are often applied to detect patterns that humans may overlook.

2. Reinforcement Learning Trading Agents

Reinforcement learning allows AI agents to learn trading strategies through trial and error. Instead of predicting prices directly, the AI learns to maximize profit by:

These models simulate thousands of trading scenarios to refine strategies.

3. On-Chain Data Analysis

Crypto markets provide a unique advantage: transparent blockchain data. AI models can analyze:

By combining on-chain analytics with market data, AI systems attempt to detect early signals of market trends.

Limitations of AI Prediction

Despite the promise, predicting financial markets — especially crypto — remains extremely difficult, even for advanced AI systems.

1. Markets Are Highly Chaotic

Crypto markets are influenced by countless unpredictable factors, including:

• Regulatory news

• Macro economic changes

• Social media sentiment

• Whale activity

Even the most advanced models struggle to incorporate sudden events that can instantly move markets.

2. Overfitting Is a Major Problem

Many AI models perform extremely well in backtests but fail in live markets. This is often due to overfitting, where a model memorizes historical data rather than learning genuine patterns.

In simple terms:
The model learns the past perfectly but fails to generalize to the future.

3. Alpha Decay

When a profitable trading strategy becomes widely used, its edge quickly disappears. AI strategies are no exception.

As more funds deploy similar models, the market adapts, and the advantage fades. This constant cycle forces traders to continuously develop new models.

4. High Competition From Institutional Quant Firms

Large hedge funds and proprietary trading firms already deploy highly sophisticated machine learning systems. Competing against these players requires massive data infrastructure, computing power, and research teams.

For most retail traders, replicating this level of sophistication is nearly impossible.

The Data Quality Problem in Crypto

One of the biggest obstacles to AI prediction in crypto markets is data quality.

Machine learning models rely heavily on large, clean datasets. Unfortunately, crypto data often contains serious issues.

1. Market Fragmentation

Crypto trading happens across hundreds of exchanges, each with different:

• Liquidity levels

• Order books

• Price discrepancies

This fragmentation makes it difficult to build unified datasets for accurate modeling.

2. Fake Volume and Wash Trading

Many smaller exchanges inflate trading volume through wash trading. If this distorted data enters a training dataset, AI models can learn misleading signals.

This leads to inaccurate predictions.

3. Limited Historical Data

Compared to traditional markets like equities or forex, crypto markets are relatively young. Many assets have only a few years of reliable historical data.

For complex machine learning models, this limited data can significantly reduce predictive accuracy.

4. Rapid Market Evolution

Crypto markets evolve faster than most financial systems. New narratives — DeFi, NFTs, AI tokens, meme coins — constantly reshape trading behavior.

A model trained on data from two years ago may already be outdated.

So… Can AI Actually Predict Crypto Markets?

The honest answer: sometimes — but not consistently.

AI can be extremely useful for:

However, fully predicting price movements remains incredibly difficult due to the chaotic and rapidly evolving nature of crypto markets.

The most successful strategies today usually combine:

In other words, AI is a powerful tool — but it’s not a magic crystal ball.

The Future of AI in Crypto Trading

While AI may not perfectly predict markets, its role in crypto trading will continue to grow.

The next generation of trading systems is already emerging, including:

• Autonomous AI trading agents

• AI-driven DeFi portfolio managers

• Real-time on-chain intelligence systems

• Cross-chain liquidity prediction models

Instead of replacing traders, AI will likely become a co-pilot for decision-making, helping traders navigate increasingly complex markets.

The hype may be loud — but the technology is still evolving.

Final Thoughts

AI has undoubtedly changed the landscape of crypto trading, offering powerful tools for analyzing massive datasets and identifying hidden patterns. However, the idea that AI can consistently predict crypto markets remains largely exaggerated.

Markets are adaptive, unpredictable, and constantly evolving — qualities that challenge even the most advanced machine learning systems.

The real opportunity lies not in blindly trusting AI predictions, but in combining human judgment with intelligent algorithms to build more resilient trading strategies.

Because in crypto, the edge rarely comes from one tool alone.

REQUEST AN ARTICLE

A huge week lies ahead on the US economic calendar with more inflation data and a Fed rate decision as markets continue to react to the war in the Middle East. 

Crypto markets are having a rare green morning during Asian trading, with most assets gaining over the past 24 hours. However, there could be more volatility in the week ahead with all eyes on the Federal Reserve meeting and what Chair Jerome Powell says about the impact of the war in Iran on inflation.

Meanwhile, President Trump plans to announce that multiple countries have agreed to form a coalition that will escort ships through the Strait of Hormuz, as fuel prices across the globe continue to increase.

Economic Events March 16 to 20

Markets will react to the US strikes on Kharg Island, an area vital to Iran’s oil industry, over the weekend, and stock futures have turned green while oil prices are back at $100 per barrel.

Wednesday is the big day for economic news, with February’s PPI Inflation report, which is unlikely to change the Fed’s hawkish stance. The US central bank meets on Wednesday, where there will be a decision on interest rates, but CME futures markets predict a 99% probability of no change.

“The Fed is going to be front and center, especially given the fact that we have seen the market push back… these rate cut expectations,” said Angelo Kourkafas, senior global investment strategist at Edward Jones.

Investors have been hoping for more rate cuts this year, which are generally bullish for stocks and crypto assets; however, those expectations have been dialed back due to fears that the surge in energy prices will push up inflation.

It will be Jerome Powell’s second-to-last meeting before his term as chair expires in May, so the next rate move may not come until Trump’s nominee Kevin Warsh takes over the helm later this year.

The rest of the week will see the Philly Fed Manufacturing Index and January New Home Sales data on Thursday.

You may also like:

“We now have the Iran war, inflation data, and a Fed meeting all in the same week,” said the Kobeissi Letter.

Key Events This Week:

1. Markets React to US Strikes on Kharg Island – Today 6 PM ET

2. February Pending Home Sales data – Tuesday

3. February PPI Inflation data – Wednesday

Advertisement

4. Fed Interest Rate Decision and Statement – Wednesday

5. Philly Fed Manufacturing Index – Thursday…

— The Kobeissi Letter (@KobeissiLetter) March 15, 2026

Crypto Market Outlook

Around $70 billion has been added to the total market capitalization over the weekend, which has climbed to $2.54 trillion on Monday morning.

Bitcoin tapped $74,000 in early Asian trading but again met resistance there and started to pull back. Ether prices continued to grind slowly higher, going past $2,200 for the first time in months.

The altcoins were generally mixed with smaller gains for Solana, Chainlink, Zcash, and Bittensor.

There’s a new debate over whether a continued pivot from Bitcoin miners to artificial intelligence could have an impact on Bitcoin security and its role as a store of value. 

While some argue that miners fleeing the network would leave it more susceptible to a “51% attack,” others argue it will simply trigger the Bitcoin network to rebalance itself as designed, making it enticing for miners again.

“AI has killed Bitcoin forever,” said crypto trader Ran Neuner on Sunday, arguing that it has become Bitcoin mining’s biggest competitor because both industries compete for electricity.

“AI is willing to pay much more for it,” he added, explaining that Bitcoin (BTC) mining revenue per megawatt is around $57 to $129, but AI data center revenue per megawatt is up to eight times higher at $200 to $500 for the same electricity, which is why miners are starting to pivot.

Earlier this month, Core Scientific secured up to $1 billion in credit for AI hosting, MARA Holdings recently filed with the SEC to signal its intent to sell some of its BTC in an AI pivot and Hut 8 signed a $7 billion AI infrastructure agreement with Google in December, argued Neuner.

Meanwhile, Cipher Mining cut its hashrate to focus on AI compute, and Bitmain cofounder Jihan Wu has stopped mining and pivoted to AI, he added.

“So if I were a miner, it wouldn’t be a tough decision. And that’s why every day more and more miners are leaving the network.” 

It sounds like a doomsday scenario for Bitcoin, but not everyone agrees. 

Bitcoin pioneer and cryptographer Adam Back argued that difficulty adjustments would only force the least efficient miners out, and profitability would improve. 

“What happens to Bitcoin is simple: tick tock, next block! Difficult adjusts downwards, the least efficient and AI switchers move out, and Bitcoin mining profitability converges to AI profitability. QED.”

“If AI outbids miners for electricity, miners just turn off until the difficulty adjusts and it’s profitable again, that’s literally how Bitcoin works,” added investor Fred Krueger.

Bitcoin energy demand is variable

However, Neuner argued that falling hashrates, which are down 14.5% since their October peak, mean that there are fewer miners to secure the network, and a higher potential for 51% attacks.

This has all happened before during bear markets, and automatic network difficulty adjustments usually compensate for it, “but this time is different because we don’t have the energy,” he said. 

Related: Crypto miners must put their Bitcoin to work to survive: Wintermute

Bitcoin ESG specialist Daniel Batten disagreed and said it was the other way around, as “the evidence tells us that AI is dependent upon Bitcoin for its expansion.”

It wasn’t all about high demand and expensive power, as Bitcoin mining can use stranded energy, act as a flexible load balancer for energy grids, and use older equipment for cheaper energy, he argued. 

One green candle to prevent AI competition doomsday

Neuner said one way to ensure AI doesn’t overshadow Bitcoin will depend on whether BTC prices go up.

“What I hope is that Bitcoin has one green candle. Maybe because of the war, maybe because of the regulation, who knows? But ultimately, if it has one green candle.” 

“If you’re watching the Bitcoin price action during this war, that’s exactly what’s happening,” he said, adding that the other scenario, where Bitcoin price continues to fall, is “pretty much a Bitcoin doomsday.”

Bitcoin has seen five monthly red candles in a row, something that hasn’t happened since the 2018 bear market. However, March is currently shaping up green with the asset gaining 8% so far this month, according to CoinGlass.

Magazine: All 21 million Bitcoin is at risk from quantum computers