Scam cases in Hong Kong fell by 2.9 per cent last year over the figure for 2024 after rising annually since 2019, while overall recorded crimes decreased by 5.9 per cent, according to police.In revealing the annual figures on Tuesday, Commissioner of Police Joe Chow Yat-ming also told lawmakers that fewer traditional street crimes were recorded, but homicide cases jumped 10-fold to 194 from 19, mostly due to the inclusion of the 168 people killed in the Wang Fuk Court blaze.“The main rise [excluding Wang Fuk Court victims] came from domestic violence, mainly due to financial problems or murder-suicides with patients with chronic illnesses,” Chow said.The force has arrested 16 people on suspicion of manslaughter in relation to the inferno in Tai Po last November.Overall, Hong Kong recorded 89,137 crimes last year, a 5.9 per cent drop from the 94,747 in 2024. Deception cases took up nearly half the number, accounting for 48.5 per cent.There were 43,212 scam cases last year, down by 2.9 per cent in 2024 and marking the first annual decline since 2019.
Scam cases fall for first time since 2019 in Hong Kong, annual figures show
South China Morning PostEN 1 min read 75% complete by Jess MaFebruary 10, 2026 at 03:51 PM

Article Content
short article 1 min
Key Entities & Roles
Key Player Opposition Context
Keywords
scam cases 100% crime rate 80% hong kong 70% annual figures 60% homicide cases 50% domestic violence 50% street crimes 40% financial problems 40%
Sentiment Analysis
Neutral Score: 0.00
Source Transparency
Source
South China Morning Post
Classification Confidence
90%
Geographic Perspective
Hong Kong
This article was automatically classified using rule-based analysis.
Topic Connections
Explore how the topics in this article connect to other news stories
Network visualization showing 21 related topics
View Full Graph scam casescrime ratehong kongannual figureshomicide casesdomestic violencestreet crimesfinancial problems
Explore Full Topic Graph Find Similar Articles
AI-PoweredDiscover articles with similar content using semantic similarity analysis.