Artificial intelligence is quietly revolutionising how we generate, distribute, and manage energy. It helps balance electricity demand in real time, spot faults before they happen, and optimise how decentralised systems like rooftop solar or batteries interact with the grid. But as AI takes on more responsibility, we have to ask: are we building resilience, or are we introducing new vulnerabilities?

The truth is, many of the systems AI depends on are deeply sensitive. They know when you’re home, how much energy you’re using, even what appliances you’re running. While this data is essential to modernising our grid, it’s also a tempting target for cyberattacks, and we’re already seeing the consequences. According to the International Energy Agency (IEA), cyberattacks on energy utilities have tripled in the last four years.

On December 23, 2015, a coordinated and sophisticated cyberattack led to a widespread power outage lasting six hours, leaving hundreds of thousands of residents in and around Kyiv without electricity. It marked one of the first known instances where hackers successfully disrupted a national grid.

These attacks aren’t just about data breaches. They can manipulate devices like smart meters or inverters and even corrupt the data that AI models use to make decisions. This tactic is known as data poisoning. That’s when the machine-learning brain behind the grid gets fed misleading information, triggering outages or massive inefficiencies. 

It’s not just about AI. The energy grid now relies on a digital supply chain that includes cloud vendors, imported chips, and third-party APIs. Every layer adds another door through which bad actors can walk. The more complex the system, the more surface area there is to attack.

Many would argue that AI’s predictive threat modelling can be a solution, but it has its own limitations. The idea of “AI snake oil,” introduced by Princeton professor Arvind Narayanan, challenges the belief that AI can solve every problem, especially in fields like cybersecurity. Predictive AI learns from past data, but threats in the real world are constantly evolving. Attackers don’t play by the same rules, and that makes it easy for them to slip through the cracks. Worse, it often flags harmless behaviour as suspicious, flooding teams with alerts and making it harder to spot real issues. In critical systems like energy, we need more than just smart tools, we need trustworthy ones.

That’s where blockchain comes in.

Blockchain gives us a way to secure the integrity of the data AI depends on. By locking in information at the point of origin with time stamps and digital signatures, we can create a tamper-proof record of every action. So when AI reroutes power or shuts off a section of the grid, we don’t just hope it made the right call. So even if AI misses something or gets it wrong, the underlying data stays trustworthy. And when AI does flag something, blockchain gives teams the context to understand what actually happened, cutting through the guesswork.

It also helps with identity and access. In a decentralised energy world, how do we know which devices are real? Who owns which EV? With decentralised ID (DID), blockchain ensures only trusted participants are allowed into the system, without exposing anyone’s private data.
There’s another strength that’s often overlooked: resilience. Centralised systems have single points of failure. Distributed ledgers don’t. If one node goes down, the others keep going. That’s the kind of architecture we need for something as critical as the power grid. Blockchain enables multi-party approvals and smart contract-based controls, so no one actor can compromise the system.

Finally, blockchain improves coordination. AI thrives on shared data. Blockchain makes that possible without compromising security. It connects the dots while protecting what matters.

At Powerledger, we believe the energy grid of the future should be:

• AI-augmented
• Decentralised
• Auditable
• Resilient by design

Blockchain isn’t just a trust layer. It’s the ledger of truth. And in a world where AI is helping decide when the lights stay on, that kind of truth isn’t optional.

Want to explore the intersection of blockchain, AI, and energy? Let’s talk.