Could a software failure cause a cascading blackout? A checklist

Software can contribute to a large blackout, but the record shows it usually does so alongside human, procedural and physical causes. The 2003 North American blackout involved failed monitoring tools. The 2015 Ukraine attack caused outages for about 225,000 customers through remote access. Neither proves a continent-wide software blackout is likely. Both justify preparing.

What does the 2003 report say?

The joint US and Canada task force report says the August 14, 2003 blackout, which the Department of Energy page says affected about 50 million customers, came from four groups of causes.

1. FirstEnergy and its reliability council failed to understand the inadequacies of FirstEnergy's system, particularly voltage instability, and did not operate with appropriate voltage criteria.
2. FirstEnergy had inadequate situational awareness. The report lists missing procedures to ensure operators knew the state of their critical monitoring tools, no effective back-up monitoring after the primary monitoring and alarming systems failed, and weak communication between computer support and operations staff.
3. FirstEnergy did not adequately manage tree growth in transmission rights of way, which caused the outage of several lines.
4. The reliability organizations did not provide effective real-time diagnostic support, with one lacking real-time data in its state estimator and another using non-real-time data.

The takeaway: software was part of the chain, mostly as monitoring and data that humans relied on, and it failed alongside vegetation management and procedures. The report is about practices, equipment and decisions, and we are not summarizing it as a software bug story.

What about deliberate attacks?

CISA's alert on the December 23, 2015 attack on Ukrainian power distribution says remote intrusions hit three regional distribution companies and affected about 225,000 customers. Operators opened circuit breakers remotely using existing remote administration tools or control system client software over VPN connections, and the companies believe the attackers had obtained legitimate credentials earlier. Malware wiped systems, serial-to-Ethernet devices were disabled, and power was restored. The alert recommends controlled remote access, network isolation, application allow listing, and contingency plans assuming the control system may be working against safe operation. The US Government attributes the campaign to Russian state actors.

That event was an outage of distribution companies, not a cascade across a grid. We found no source in this research that quantifies the probability of a software-driven cascading blackout, so we do not offer a number.

What makes cascades possible?

The 2003 report shows how local failures spread when operators cannot see what is happening. Common ingredients: tightly coupled systems, loss of visibility, delayed decisions and weak coordination between organizations. Software changes the speed and the visibility, which is why monitoring tools and their backups matter.

Preparedness checklist for an organization

1. Know your dependency on power. List the systems and sites that stop without grid power, and for how long you can run on backup.
2. Test backup power. Run the generators and batteries under load on a schedule.
3. Plan manual operation. Write and rehearse how to run critical processes with screens down or untrusted.
4. Back up monitoring. The 2003 record points to the cost of losing the tools that show system state. Keep an independent view.
5. Limit remote access. Time-limited, logged, with strong authentication, and no persistent vendor connections into control networks.
6. Keep offline copies. Configurations, procedures and contact lists in paper or offline storage.
7. Know who to call. Maintain contacts for your utility, suppliers and emergency services.
8. Practice together. Run a drill with operations, IT and facilities in the same room.

What to do

1. Pick one critical site and run through the checklist in an afternoon.
2. Fix the cheapest gap this week, usually contact lists and printed procedures.
3. Schedule a backup power test within the quarter.
4. Review remote access paths and remove any you cannot justify.
5. Repeat for the next site.

Quick check

Sources

• US-Canada Power System Outage Task Force: August 14th Blackout, Causes and Recommendations
• US Department of Energy: August 2003 Blackout
• CISA: Cyber-attack against Ukrainian critical infrastructure

Last reviewed: 18 October 2026.