HFX-Ipsos Threat Index Finds Disinformation, Hacking Seen as Top Threats

Disinformation and hacking remain tied as the top perceived threats for the second year in a row.

A new Ipsos poll of more than 23,000 people across 30 countries – the HFX-Ipsos Threat Index – conducted by Ipsos for the Halifax International Security Forum finds just over two in three think some person, organization, or country deliberately spreading disinformation to influence public opinion (stable year-over-year at 77%) in the next twelve months is a real threat.

Worry about disinformation still looms large

We added disinformation to the poll last year as the issue dominated headlines in the run-up to the U.S. presidential election on November 5, 2024. This year’s polling, conducted on Ipsos’ Global Advisor online platform between Sept. 19 and Oct. 3, 2025, finds worry about disinformation post-election has eased a bit among Americans to 82% (-six percentage points). While concern about the issue in Canada, which held a federal election earlier this year, remains relatively steady (79%, +two pts).

People in Thailand (87%, +two pts) are currently the most worried out of 30 countries about disinformation. People in India, which also held general elections in 2024, are now the least concerned (57%, -16 pts). 

Threat of hacking ties for top spot

Cyberhacking has consistently been a leading concern over the past decade in our survey and ties disinformation as the No. 1 threat this year, with 77% (stable) on average globally saying that the threat of some person, organization or country hacking into either their public, private or personal information system for fraudulent or espionage purposes in the next twelve months is a real threat.

Across all 30 countries, a nuclear/biological or chemical attack taking place somewhere in the world is currently perceived as a real threat by 72% (no change), followed by natural disasters (67%, -three pts), a terrorist attack (66%, no change), the personal safety/security of individuals/their family members being violated (62%, no change), an artificial intelligence-based defense system becoming a threat to humanity by breaking free of human control (62%, +two pts), a violent conflict breaking out between ethnic/minority groups in their country (61%, stable), a major health epidemic (60%, -two pts) and armed conflict with another country (52%, +two pts).

Since 2016, on average across 20 countries*, hacking for fraudulent or espionage purposes has been seen as the threat most widely found to be real except for in 2020 (the year a global pandemic was declared and a major health epidemic topped the list) and 2022 (the year Russia invaded Ukraine when hacking tied with fears about a nuclear/chemical attack). Hacking has now tied with disinformation, added in 2024 amid a historic election year, as the top threat for the second year running.

The proportion of citizens listing cyberattacks as a real threat significantly declined in India (-15 pts to 58%) in 2025. Meanwhile, concern remains just above the global average in Canada at 79% (stable) and America (82%, -two pts).

Nuclear/biological/chemical attack concern eases a bit

Concern about a nuclear, biological, or chemical attack happening somewhere in the world rose nine points to 74%, on average across 20 countries*, in the wake of the full-scale invasion of Ukraine in February 2022. Despite easing to 69% in 2025 as the conflict nears the four-year mark, worry is still slightly higher than it was in 2021 when 65% considered it a real threat. Though it’s important to note 2025 polling was done before U.S. President Donald Trump claimed earlier this month that America could “destroy the world 150 times” with nuclear weapons and before Russian President Vladimir Putin announced Russia had tested a new weapon.

Worry about a possible attack remains widespread and is seen as a real threat by majorities in every country surveyed in 2025, ranging from 56% (-two pts) in France to 89% (+three pts) in Thailand. 

People in Brazil (-11 pts to 61%) are significantly less worried about a nuclear, biological, or chemical attack than they were last year, while concern rose eight points among Britons to 80%. Worry in the U.S. remained slightly higher than the global average at 76% (-two pts) and dipped a bit in Canada (-four pts to 70%).

Concern about an epidemic remains elevated

The coronavirus crisis is now firmly in the rearview mirror, but scars remain.  

The World Health Organization declared a global pandemic on March 11, 2020, and declared an end to the COVID-19 global public health emergency more than three years later on May 5, 2023.

The proportion of people worried about a major health epidemic breaking out in their country has now levelled out around the 60% mark. While that’s well below the 79%, on average across 20 countries*, who considered an epidemic a real threat in 2020 during Year One of the pandemic it’s still above the 51% who thought the same in 2019 pre-pandemic.

Lack of confidence in government and its agencies to deal with threats

Throughout the world, many citizens continue to be skeptical that “the appropriate levels of security and protection could be provided by [their] government or its agencies could respond effectively” to perceived real threats. 

In 2025, 77% say some person, organization or country hacking into their information system for fraudulent/espionage purposes is a real threat, yet only 47% express confidence in their government’s ability to respond to cyberhacking in their country — a difference of 30 points.

There are similarly wide gaps on a few other issues. More than two-thirds (69% on average across 30 countries) view a nuclear/biological/chemical attack taking place somewhere in the world as a real threat, yet only 40% express confidence in their government’s ability to respond if a nuclear attack took place — a difference of 29 points. And while the majority (a 30-country average of 77%) feel some person, organization, or country deliberately spreading disinformation to influence public opinion is a real threat, half (49%) are somewhat/very confident their government could respond to the threat effectively — a gap of 28 points.

Optimism falls to new low, fear stabilizes

Modern life is full of threats online and in real life.

And a strong majority (staying stable at 79% on average across 20 countries*) continue to think the world became more dangerous in the past year. But, the proportion thinking things are getting increasingly dangerous is down seven points from a high of 86% in 2022. 

In 2025, people in the Netherlands (88%, +two pts) are the most likely to say the world became more dangerous while those in India are the least likely (67%, -six pts). About eight in 10 Canadians (82%, no change) believe the world became more dangerous, while three-quarters of Americans (75%, -six pts) now think this. 
At the same time, many are feeling pretty pessimistic. 

The proportion thinking the world is getting better continues its slide to a new low of 33%, on average across 20 countries*, and is 14 points lower than in 2017 (47%) when we first asked this question. 

The Japanese are feeling the most glum, with a mere 16% (-four pts versus 2024) thinking the world is getting better while those in India (65%, -three pts) are the most optimistic. Those in the U.S. (stable at 32%) and Canada (23%, -six pts) remain down with less than one-third in both countries agreeing that more things are getting better in the world these days than are getting worse.

*Change for 20-country average is based on the countries which were in all editions of the Ipsos for the Halifax Security Forum report over the past decade.
**BRIC (Brazil, Russia, India, China) and CRINK (China, Russia, Iran, North Korea) countries.

About the Study

These are the results of a 30-country survey conducted by Ipsos on its Global Advisor online platform and, in India, on its IndiaBus platform, between Friday, September 19, and Friday, October 3, 2025. For this survey, Ipsos interviewed a total of 23,586 adults aged 18 years and older in India, 18-74 in Canada, Republic of Ireland, Malaysia, New Zealand, South Africa, Türkiye, and the United States, 20-74 in Thailand, 21-74 in Indonesia and Singapore, and 16-74 in all other countries. 

The sample consists of approximately 2,000 individuals in Japan, 1,000 individuals each in Australia, Belgium, Brazil, Canada, France, Germany, Great Britain, Italy, New Zealand, Spain, and the U.S., and 500 individuals each in Argentina, Chile, Colombia, Hungary, Indonesia, Ireland, Malaysia, Mexico, the Netherlands, Peru, Poland, Singapore, South Africa, South Korea, Sweden, Thailand, and Türkiye. The sample in India consists of approximately 2,200 individuals, of whom approximately 1,800 were interviewed face-to-face and 400 were interviewed online. 

Samples in Argentina, Australia, Belgium, Canada, France, Germany, Great Britain, Hungary, Italy, Japan, the Netherlands, New Zealand, Poland, South Korea, Spain, Sweden, and the U.S. can be considered representative of their general adult populations under the age of 75. Samples in Brazil, Chile, Colombia, Indonesia, Ireland, Malaysia, Mexico, Peru, Singapore, South Africa, Thailand, and Türkiye are more urban, more educated, and/or more affluent than the general population. The survey results for these countries should be viewed as reflecting the views of the more “connected” segment of their population.

India’s sample represents a large subset of its urban population — social economic classes A, B and C in metros and tier 1-3 town classes across all four zones. 

The data is weighted so that the composition of each country’s sample best reflects the demographic profile of the adult population according to the most recent census data. “The Global Country Average” reflects the average result for all the countries and markets in which the survey was conducted. It has not been adjusted to the population size of each country or market and is not intended to suggest a total result.

When percentages do not sum up to 100 or the ‘difference’ appears to be +/-1 percentage point more/less than the actual result, this may be due to rounding, multiple responses, or the exclusion of “don't know” or not stated responses.

The precision of Ipsos online polls is calculated using a credibility interval with a poll where N=1,000 being accurate to +/- 3.5 percentage points and of where N=500 being accurate to +/- 5.0 percentage points. For more information on Ipsos' use of credibility intervals, please visit the Ipsos website.
The publication of these findings abides by local rules and regulations.