How AI is Disrupting Politics

When thinking of AI and its involvement in politics, it’s fair to assume the stereotypes are not positive. AI’s influence in politics over the past few years has come under serious fire. Facebook’s algorithms, for example, have been blamed for spreading of misinformation, which heavily impacted major political events such as Brexit and the 2016 US election.

However, AI’s future within politics isn’t all bad. As the technology advances, it has demonstrated great potential to improve the political landscape it had previously tainted. Here’s how:

It can help us form better laws

Currently, there are more than 8,550 bills and resolutions before Congress. However, only about 7 percent of them actually pass. With such a low pass rate, it can be hard for those in Congress to know where their time is best spent. If chosen incorrectly, a lawmaker could spend countless hours on a bill that never passes, resulting in nothing more than wasted time and energy.

Even more, major bills are typically more than 1,000 pages long. The pure amount of information is almost impossible for humans to make sense of. As a result, lawmakers are forced to vote without a deep understanding of the bill they are passing. This was recently brought to light with the tax reform bill at the end of December, where many lawmakers had just a few hours to read the new version. An “impossible” task, to read so much in such as short space of time, as US senator, Chuck Schumer, stated himself.

So how can AI help improve this? The good news is that lawmakers no longer have to attempt this impossible task, because AI can read it for them. Consequently, pinpointing the essential information means lawmakers spend less time reading and more time evaluating their options and outcomes.

At our own company, for example, we’ve tapped into our team’s background in building natural language processing engines for the National Security Agency and the Department of Defense to track and analyze an ontology of over 50 million key phrases that lawmakers have used. This, however, is just one example of the current abilities of AI to enhance government processes.

Algorithms can also quickly scan text and process huge amounts of historical voting data, as well as data on the current political climate, to allow us to predict the likelihood of a bill passing. In turn, this enables lawmakers to focus their energy — and thus, work on bills that matter most.

It can allow smaller businesses to be heard

If your business isn’t concerned by the decisions made by Congress, then it should be, because it is affecting your earnings. A previous McKinsey reported stated, “The business value at stake from government and regulatory intervention is huge: about 30 percent of earnings for companies in most industries.”

Large businesses are fully aware of the impact lawmaking can have on their industries. This is reflected in the billion of dollars pumped into lobbying to influence the outcome of laws in their favor.

However, many smaller business do not have colossal bank accounts to support their best interests, yet still have a voice which deserves hearing. This point was emphasised by the current controversy surrounding Deferred Action for Childhood Arrivals (DACA). Small businesses, alongside big businesses, have expressed a strong desire to maintain DACA, for moral and economic reasons, with some claiming that DACA deportations could cost the US more than $400 billion.

In many ways the issues faced by small business are representative of the same issues faced by lawmakers: too much information and not enough time. So deciding which laws to focus on or who to engage with can be a lengthy and, at times, fruitless process. However, AI has an opportunity to help.

Thanks to technology, just like lawmakers, smaller businesses can use AI to understand issues that affect them and which legislator can help them with the causes they care about while allowing them to stay up to date on legislation.

It will rethink public sector work

For AI and politics, providing clearer summaries and great access to information is only the beginning. Due to its ability to automate simple tasks, we are likely to see this technology free up jobs, shifting attention to more essential areas, and enabling the workforce to be utilized in the most efficient way possible.

According to Deloitte’s AI Augmented Government, large government job losses from AI are unlikely. However, “cognitive technologies will change the nature of many jobs — both what gets done and how workers go about doing it — freeing up to one quarter of many workers’ time to focus on other activities.”

Moreover, millions of working hours will slowly become automated, carried out by computers which already have the capability to complete these tasks in their entirety. The Deloitte report estimates that, at the low end of the spectrum, automation could save 96.7 million federal hours annually. This could lead to potential savings of $3.3 billion. Alternatively, at the high end, 1.2 billion hours could be saved with potential annual savings reaching $41.1 billion.

One example of this is the government’s use of AI for real-time tracking and reporting for disease surveillance. The US Centers for Disease Control and Prevention has adopted this technology to improve poliovirus tracking and the reporting process with an AI tool, classifying virus types and separating disease reports into related groups. Other laborious tasks can also be reduced using AI, such as dealing with paperwork burdens by processing text or daunting Medicare backlogs, a monumental task which officials have claimed will not be achieved by 2021, the deadline set by Health and Human Services (HHS).

Technology has often pioneered new ways to make our lives easier and more efficient, and despite a shaky start into the world of politics, AI has the potential to do the very same thing for government. In doing so, governments can alleviate themselves of tasks which are both time and resource heavy, streamlining processes, and allowing all of us to do more with less.