M+E Daily

New Knowledge CEO: AI, Machine Learning Can Be Used to Combat Manipulative Campaigns

As we continue to hear how the Russian government and its troll armies used paid advertisements and viral propaganda to interfere in the 2016 U.S. presidential election, media and entertainment (M&E) companies must now combat new adversaries who have been using the same tactics to attack brands and manipulate the media, according to Jonathon Morgan, CEO of cybersecurity company New Knowledge and founder of Data for Democracy.

There are several steps that M&E companies can take to combat “large-scale, manipulative campaigns” against their brands and artificial intelligence (AI) and machine learning (ML) can help significantly, he said July 24 during the closing keynote session called “Defending Against Disinformation and Social Media Manipulation” at Content Protection Summit East, part of the Media & Entertainment (M&E) Day at the Microsoft Conference Center.

The 2018 M&E Day also included Smart Content Summit East and Entertainment Production in the Cloud (EPIC) conference tracks, providing M&E technology teams valuable insights into the creation, production, distribution, security and analysis of content.

It’s important for companies to discover manipulative campaigns against their brands before their customers do, Morgan told the summit.

To do that, for starters, companies need to monitor open and closed spaces of known “bad actors” including 4chan, private Discord channels, subreddits and certain Facebook groups, according to Morgan.

What’s especially dangerous about such campaigns to M&E companies is that – especially in comparison to the manpower and money Russia used to interfere with the 2016 U.S. presidential election – it doesn’t take much to harm an M&E brand or piece of content, he pointed out.

“The kind of people who are undermining entertainment – they don’t need those types of resources because the surface area is so much smaller,” Morgan said, adding: “You don’t need to do very much work to manipulate a conversation about a movie release. So, with a couple of hundred dollars and a thousand of your friends you can completely undermine a conversation about something at this scale. It’s cheap. It’s easy.”

Companies must react quickly to such campaigns before they break, and automated, real-time early warning alerts distributed via email, text or team collaboration applications including Slack can help, he said. Machine learning signals are valuable because they can detect anomalous changes in community behavior, according to Morgan.

Next is to “inoculate your audience” to “neutralize the threat,” he told attendees.

But “the long-term solution has got to be about partnerships and building alliances” including collective industry pressure on review aggregators and social media platforms and legislation, he said. Preparation for the next generation of spear-phishing is also needed, according to Morgan.

“Social media manipulation is widespread and systemic,” he said, pointing to data showing 60 million Facebook accounts are fake, 48 million Twitter accounts are fake, and 126 million people saw election propaganda in 2016.

New Knowledge uses AI, among other technologies, to help businesses monitor and defend against the viral spread of intentionally damaging misinformation, detecting false narratives before they can make an impact, according to the company.

The company used AI and ML to identify extremism on social media platforms including Facebook, Instagram and Twitter, Morgan noted during his keynote. ML was used to identify the most polarizing, extremist content in images and text, he said. That allowed New Knowledge to target the accounts and conversations that were most influential in manipulating Americans on divisive social issues including gun rights and race, he noted, adding the company was able to measure the language being used and quantify their influence on the larger social media community. For example, the company’s platform computed how suspected Russian-linked accounts described the Jewish population, spread anti-Semitic propaganda and shaped the language of other social media users, he pointed out.

The 2018 Media & Entertainment Day was presented by Microsoft, with sponsorship from IBM Watson Media, Amazon Web Services, IBM, LiveTiles, Microsoft Azure, NAGRA, NeuLion, Ooyala, EIDR, GrayMeta, MarkLogic, Qumulo, Avid, Cloudian, SoftServe and TiVo. The event was produced by the Media & Entertainment Services Alliance (MESA), the Content Delivery & Security Association (CDSA), the Hollywood IT Society (HITS) and the Smart Content Council.

Click here for audio of Morgan’s presentation or here for the slide deck.