AI is a Pandora’s box that will completely reshape the rules of war
Finally, the impending dangers of Artificial Intelligence (AI) are making the headlines. When the “Godfather of AI” Geoffrey Hinton quits Google – speaking out about the risks it poses to human civilisation – we should listen.
Likewise, when an AI-controlled unmanned aerial vehicle “killed” its human handler in a simulated test, eliminating the operator for daring to interfere with its mission objectives, we should worry. But when security minister Tom Tugendhat (a rare round peg in a round hole) warns us that AI is moving too fast for even the “finest clerks” in Parliament to regulate – Westminster, we have a problem.
Parliamentary collective wisdom is probably low. My immediate apologies to any MP with a PhD in machine learning! But such is the gargantuan impact this Pandora’s box will have on all areas of our lives, this must change. I have already called for a new ministerial post for artificial intelligence.
We have less than two years before AI becomes a security problem we can no longer control
AI is the ultimate force multiplier with a capacity to totally reshape the rules of war. Of course technological innovation has for centuries advanced the character of conflict. Henry V was the first to harness the longbow at Agincourt, the use of the Gatling Gun in the United States Civil War, tanks at the Battle of Cambrai, aircraft carriers at Pearl Harbor, and of course, nuclear bombs during the Second World War all added new dimensions to warfare.
Of these examples, only the last was deemed so existentially dangerous to prompt US president Dwight D. Eisenhower to call for the immediate verifiable regulation of nuclear power in his 1953 Atoms for Peace speech. We must now do the same with AI. Simply put, machine learning and autonomy is about to alter the battlefield existentially and whoever gets there first will have a critical advantage.
War fighting is about lethality – how and when individual military components are orchestrated to maximise concentration of force to overcome an adversary. In removing the human from both operating those components and the decision-making process, efficiency will jump exponentially.
AI will remove the “fog of war”. Masses of data regarding enemy strengths, weaknesses, and intensions in contrast to individual dispositions will be assessed instantly. Hesitance, human error, and misguided heroism will all be removed from command and control. Utilising swarm unmanned systems will allow for instantaneous smart “sacrifices” to be calculated, the timing of which could be time critical in winning a battle.
If we outsource decision making to a machine capable of learning as it fights, then vital hours will be saved as confirmatory orders are no longer passed up and down the chain of command.
Even in the grey zone, large scale disinformation could be actively propagated across the internet far more efficiently. AI can instantly monitor and adjust its messaging to maximise political discord, even to the point where targeted communities would struggle to know what is true anymore.
By any measure our world is becoming more dangerous not less. We are conscious of a demise in global governance, the retreat from globalisation and the prospect of economic battles over critical technologies and natural resources. And we opine the ever-obvious limits of the United Nations to hold errant nations to account.
Our world is not at war, but we are certainly not at peace. Demands placed on our security services are only increasing. Yet despite conducting two full reviews of our defence posture over the last two years, which confirm a deteriorating threat picture, our defence budget has not shifted from its peacetime setting of 2.2 per cent GDP.
Two decades after Facebook came into our lives, we are finally introducing rules to limit the harms caused by social media. AI technology will no doubt be a force for good in our lives. But without the right regulations, we have less than two years before AI becomes a security problem we can no longer control.
Tobias Ellwood, Conservative MP for Bournemouth East and chair of the Defence Committee
Get the inside track on what MPs and Peers are talking about. Sign up to The House's morning email for the latest insight and reaction from Parliamentarians, policy-makers and organisations.