AI At The Crossroads: The Dual Role In Spreading And Combatting Narrative Attacks
By Naushad UzZaman
Artificial intelligence can spread narrative attacks, but it can also combat it. Blackbird.AI's CTO explains why AI experts must work now to reduce polarization, build trust, and expand opportunities to curb false narratives.
As the CTO and co-founder of Blackbird.AI, a company dedicated to fighting narrative attacks caused by narrative attacks, I spend a lot of time thinking about artificial intelligence (AI) and its impact on society. There’s been a lot of hype and speculation lately about AI, especially with the recent excitement around generative AI tools like ChatGPT, Claude, Gemini, and others. People are understandably excited but apprehensive, wondering if AI will soon reach human-level intelligence, a threshold known as artificial general intelligence (AGI).
While AGI is a fascinating topic to ponder and presents many long-term questions for humanity, there are more urgent AI-related threats we need to focus on first. Narrative attacks, powered by today’s AI tools, poses an immediate danger to individuals, organizations, and democracy. We must avoid getting distracted by speculative, far-off scenarios before addressing AI-enhanced narrative attack’s clear and present challenges.
Co-founder and CTO Naushad UzZaman explains how Blackbird.AI blends artificial intelligence with human analysts to protect organizations from narrative attacks that cause financial and reputational harm.
It’s important first to clarify what AI is and how it’s already integrated into our daily lives. At its core, AI is about creating machines that can perceive, learn, and act in ways that mimic human intelligence. Some of the earliest visions for AI imagined that we could create machines that think like humans in a matter of years or decades. That hasn’t panned out yet, and we still don’t know if it ever will. But what has happened is that AI has become exceptionally good at assisting humans and augmenting our capabilities.
Think about all the AI systems you interact with regularly, probably without realizing it. When you shop on Amazon or watch Netflix, AI algorithms make personalized recommendations. Google and Facebook use AI to optimize which ads to show you. The newsfeeds and content you see on social media are all curated and customized using AI. Voice assistants, fraud detection, facial recognition—AI is everywhere.
More recently, the incredible fluency of large language models like GPT-4 has fueled speculation that AGI may be just around the corner. These models can engage in human-like dialog, write coherent articles and stories, answer questions, and write code. In many cases, you’d need help distinguishing their output from something written by a human.
But here’s the thing – you don’t need AGI to cause massive real-world harm with AI. Today’s AI technologies are already being weaponized at scale to spread narrative attacks, manipulate populations, and undermine trust in social institutions. As the World Economic Forum has stated, the global narrative attacks crisis is humanity’s number one short-term threat. At Blackbird.AI, we see this playing out every day.
Nation states, extremist groups, and bad actors are leveraging AI to automate influence operations at an unprecedented speed and scale. They harness AI to identify the most divisive narratives, craft authentic-seeming messaging, and deploy armies of human-like bots to spread this content far and wide. With just a single prompt, today’s AI tools can generate an entire narrative attack campaign that gets further amplified by bot networks.
The result is an information ecosystem so polluted with misleading propaganda that people no longer know what or who to believe. Overwhelmed by an unending deluge of conflicting and emotionally charged content, many are simply tuning out or losing trust in all institutions and authority. When people can no longer agree on basic facts, democracy itself is threatened. You don’t need AGI for society to tear itself apart – today’s AI-powered narrative attacks is already pushing us to the brink.
At Blackbird.AI, we’re using AI to fight back against this onslaught of narrative attacks. Our Constellation Narrative Intelligence Platform leverages multiple AI models to detect and map narratives as they emerge. We look at the elements of narrative attacks like a constellation – the narratives being pushed (the “what”), the actors behind them (the “who”), the tactics used to spread them (the “how”), and the malign goals driving them (the “why”).
Some of the key AI capabilities we use at Blackbird.AI include:
- Narrative detection: We automatically identify and cluster the overarching storylines and messaging in a data set. This allows us to surface emerging narratives early before they reach critical mass.
- Cohort analysis: Profiling the actors promoting a narrative to understand their demographics, ideological leanings, and motives. Are they right-leaning, left-leaning, anti-vaccine, pro-Russia, etc.? Knowing who is behind a narrative provides crucial context.
- Manipulation detection: We analyze behavioral signals to assess whether bots and coordinated networks artificially amplify a narrative. We look for anomalous patterns like posting frequency, copied content, suspicious account creation dates, etc.
- Network mapping: Visualizing how narratives flow and spread between accounts over time. This lets us distinguish organic virality from coordinated promotion by inauthentic networks.
By integrating all these AI-derived insights, Constellation gives our customers a real-time situational awareness of threats to their brand so they can respond quickly and effectively. We enable a proactive rather than reactive defense against narrative attacks.
However, exposing narrative attacks is only half the battle. We must also equip people with the tools to think critically about the deluge of online content. That’s why we built Compass by Blackbird.AI, an AI-powered application that allows anyone to easily “context check” claims they see on social media or in the news.
Compass searches the live web for authoritative information on a topic and summarizes the key context and facts needed to assess the claim’s credibility. The user doesn’t need to spend hours chasing down sources and evidence—Compass does the heavy lifting and delivers a concise, reliable briefing. It’s like a personal research assistant to help navigate today’s challenging information environment.
Importantly, we see Compass not as a substitute for critical thinking but as a tool to enhance people’s judgment and agency. By making it easy to access reputable information that may otherwise get drowned out by noise and sensationalism, we aim to empower people to reach their well-informed conclusions. In a world of information overload and rampant narrative attacks, tools like Compass will be increasingly essential for an engaged and discerning citizenry.
While AGI may be a fascinating long-term prospect, our priority should be addressing the immediate harms posed by narrow AI today, especially in narrative attacks. We need a societal mobilization of resources, policies, and innovations to safeguard the integrity of the information ecosystem on which democracy depends. This is an all-hands-on-deck moment requiring courageous leadership from government, industry, and civil society.
Some key steps I believe we must take:
- Dramatically increase public awareness of online manipulation, including teaching media literacy, tech literacy, and critical thinking skills from an early age. For example, most people still think that virtuality is random and that social media posts gain popularity by chance. Many do, but so many are manipulated for a purpose. Good actors, bad actors, and everyone in between are vying for attention. Just as we teach kids to be safe drivers, we must teach them to be safe information consumers.
- Pass thoughtful regulations that increase transparency and accountability in the tech platforms, ad networks, and other intermediaries that are vectors for narrative attacks. We cannot allow a hands-off, profit-driven approach anymore.
- Support and scale independent research, fact-checking, and counter-narrative attack solutions. We need a robust ecosystem of credible voices with cutting-edge AI tools to detect and refute harmful narratives in real-time.
- Address the underlying social, economic, and political conditions that make people vulnerable to narrative attacks in the first place. We must work to reduce polarization, build trust, and expand economic opportunity so people feel less isolated, aggrieved, and willing to believe convenient lies. Compass by Blackbird.AI, for example, is a tool that can move the needle in a positive direction by rapidly providing context, nuanced details, and even citation links about some of the most complex and controversial topics.
The age of AI is here, and it will bring many disruptive changes, both positive and negative. We’ll address job displacement through re-skilling programs, educational reforms, and policies like universal basic income. But if we fail to handle AI-powered narrative attacks immediately, we may find ourselves in a “post-truth” world where none of our social institutions can function.
Malign actors already use today’s AI to hack our minds at scale and pit us against each other. We cannot sleepwalk into this dystopia. It’s time to come together and fight back with our own AI innovations grounded in defending truth and democracy. At Blackbird.AI, this mission animates everything we do. The stakes could not be higher.
While we may not have all the answers yet, one thing is clear to me – addressing the narrative attack crisis should be the top priority of our time. Nothing less than the future of human agency and self-determination is on the line.
Let’s get to work!