If executives are not paying attention to alternative data – specifically AI-powered analysis of social media and online news – they are allowing a giant blind spot to remain in their risk-analysis vision.

Executives are being told, from every institutional direction, that the pace of country-level surprise and risk has accelerated. But many are still relying on the same sources they read five years ago: country risk scores , IMF Article IV reviews and OECD classifications that that update quarterly at best and annually at worst.

At the other extreme are news headlines: Informative in a broad way, but immediate, visceral and inflationary. Nothing can make you a prisoner of the moment like the 24/7 news cycle. Whipsawing between these slow-thinking and fast-moving sources means that executives are forced to make 21st-century capital decisions on a 20 th -century information diet.

The space between those extremes is where alternative data lives and, within it, the continuous monitoring of social media and structured online news, in local languages at national and even sub-national resolution. This is the single most underused input in the corporate risk stack.

Alternative data is an umbrella term for nontraditional, non-financial information that reveals exposure that conventional instruments miss. It can include satellite imagery, shipping telemetry, payment flows, corporate registry changes and regulatory filings harvested at scale. The number of alternative-data providers has grown more than twenty-fold over the past three decades, from roughly 20 in 1990 to more than 400 by 2023.

But for executives trying to understand country-level risk specifically, one source towers over the others in value and remains the most persistently underused: the combination of social media monitoring and structured online news analysis across local-language sources.

Expect Push Back From the C-Suite

Analysts who urge their executives to consider this approach should expect push back. Social media is still dismissed by too many boardrooms as marketing's sandbox rather than as a serious forecasting enterprise. That impression is at least a decade behind both the times and technology. The current generation of tools ingests tens of millions of posts per day across dozens of languages, applies sentiment analysis calibrated to regional idiom and dialect, maps influence networks to separate organic discourse from coordinated manipulation and cross-references against structured feeds from thousands of local-language news outlets most Western executives have never heard of. What comes out the other side is not chatter. It is a continuously updated picture of what is being said inside a country, by whom, at what reach, and in which direction the current is running.

To show an example of the scale that is possible, one of the data-based index products my firm produces analyzed 10.2 trillion (with a “t”) pieces of online and social media data referencing 190 countries across the span of an entire year. It produced a detailed sentiment analysis of what people in the U.S. think about each country’s politics and culture.

Consider the categories of country-level risk this kind of data – and analytical power – can expose that a traditional risk score cannot.

Right Now, Not Six Months From Now

With social media, political stability registers as a measurable rise in anti-government sentiment in secondary cities right now , not as a downgrade six months after the fact. Academic research has consistently found that social media signals can meaningfully predict offline protest activity , and the University of Notre Dame’s Kroc Institute has built AI systems that monitor social media ecosystems at scale to identify trends signaling potential outbreaks of political violence.

My first exposure to this early-warning phenomenon came in 2016, when my then-Twitter feed began sending me signals about something odd happening in Turkey hours before the mainstream media reported that a coup attempt against President Erdogan was underway.

For a company doing business in a foreign country, commercial risk registers as rumors moving through domestic business networks days or weeks before a policy change hits the official gazette. A minister's political capital registers as the rising or falling frequency with which his party’s news media bother to defend him. Security trends emerge as geolocated spikes in specific incident types — strike activity, cargo theft, kidnapping chatter — often long before Western embassies revise their travel advisories.

And then there is the variable that most directly affects American companies: U.S.-directed sentiment. Is popular opinion in a particular foreign market moving toward the United States, or dangerously away? This governs whether an American brand is about to become an asset or a liability. Smartly architected social and news monitoring captures this variable directly, and in near real time.

Think of it as having on-the-ground intelligence sources everywhere you do business.

Build the Dashboard Your Risk Committee Doesn't Yet Have

The data, platforms and the AI models to do this are all commercially available. The gap is assembly: very few firms have integrated the pieces into a product that a risk committee can actually use.

The right tool is a Country Exposure Intelligence dashboard. Powered by an AI engine that ingests local-language social media, regional news, regulatory filings and NGO reporting from every market in the firm’s footprint, it runs sentiment and narrative analysis calibrated to each country. It tracks week-over-week deltas on political stability, regulatory posture, U.S.-directed sentiment, security incidents and the standing of in-country counter parties.

What’s the role of AI? It can spot narrative patterns that usually come ahead of crisis-level events — a simmering protest, a suddenly vulnerable minister, chatter about a coming regulatory crackdown. AI can flag them to human analysts before they show up in conventional data.

Let’s be clear about the most important aspect of this: Humans still make the call. Social media is full of incorrect and intentionally deceiving information. Smart, experienced risk analysts are still needed to divide the wheat from the chaff. This tool helps them, and on a timeline that matches how capital decisions actually move. The firms that will outperform over the next decade of cross-border investment are the ones that stop treating social and online-news data as the plaything of the comms team and start treating it as a standing input to the risk committee.