Trends in Spam Products and Exploits Geoff Hulten

  • Slides: 10
Download presentation
Trends in Spam Products and Exploits Geoff Hulten, Anthony Penta, Gopalakrishnan Seshadrinathan, and Manav

Trends in Spam Products and Exploits Geoff Hulten, Anthony Penta, Gopalakrishnan Seshadrinathan, and Manav Mishra Microsoft Safety Technology and Strategy Group, MSN

Data • Hotmail Feedback Loop – Volunteers classify random samples of their mail as

Data • Hotmail Feedback Loop – Volunteers classify random samples of their mail as spam or good – Tens of thousands of hand classified messages per day – Large ‘unbiased’ sample of spam • Additional analysis on two sets of spam – About a year apart – Type of product being sold – Types of spammer tricks present – Trends in spammer activity

Trends in Spam Products • Types of Products – – – – – ‘Non-graphic’

Trends in Spam Products • Types of Products – – – – – ‘Non-graphic’ Sexual Products Insurance Rx / Herbal Financial Travel / Casino Scams Newsletters Other Spam Porn/Sex Graphic Dubious Products

Most Popular Spam Product • ‘Non-graphic’ sexual products are big winners • Doubled over

Most Popular Spam Product • ‘Non-graphic’ sexual products are big winners • Doubled over the past year – 17% of spam in 2003 – 34% of spam in 2004 • About triple the volume of the next most popular product categories – Financial products – 13% – Online Prescriptions – 10% – Dubious Products – 10%

Products on the Way Out • Graphic Porn – Spam with a graphic image

Products on the Way Out • Graphic Porn – Spam with a graphic image in it – 13% in 2003, 7% in 2004 • Dubious Products – 20% in 2003, 10% in 2004

Trends in Spam Exploits • Types of Exploits – – – – Word Obscuring

Trends in Spam Exploits • Types of Exploits – – – – Word Obscuring URL Spamming Domain Spoofing Token Breaking MIME Attacks Text Chaff URL Obscuring Character Encoding • Exploits increasing rapidly – Average spam in 2003 had 1. 33 exploits – Average spam in 2004 had 1. 73 exploits

Fastest Growing Exploits • Word Obscuring – Putting words in images, misspelling, etc. –

Fastest Growing Exploits • Word Obscuring – Putting words in images, misspelling, etc. – Designed to fool content based spam filters – 4% in 2003, 20% in 2004 • URL Chaffing – Adding ‘good’ URLs to spam – Not seen in 2003, present in 10% of spam in 2004 • Spammers are reacting to filtering!

Who is Working Hardest? • ‘Non-graphic’ Sexual Products – Averages 2. 44 exploits per

Who is Working Hardest? • ‘Non-graphic’ Sexual Products – Averages 2. 44 exploits per message • Prescription drugs online – Averages 2. 12 exploits per message

Selling Porn… With a Conscience? • Spam with graphic porn – Averaged 2. 16

Selling Porn… With a Conscience? • Spam with graphic porn – Averaged 2. 16 exploits in 2003 – Averaged 1. 33 exploits in 2004 • The only category of product with a significant decrease in the average number of exploits over the study period

Summary • Top Products in Spam – ‘Non-graphic’ sexual products – Prescription medicine –

Summary • Top Products in Spam – ‘Non-graphic’ sexual products – Prescription medicine – Financial services – Dubious products • Spammers Adaptive and Working Hard to Get Spam into Inboxes – They are finding effective new tricks – They are putting more work into each spam