David Harley, the company's senior research fellow, says that, as the year progresses, we can expect to see more cross-platform threats using operating system-independent vectors like Java.
"Windows will remain the main target because it has by far the most users. And while there won't be a big shift towards specific targeting of other operating systems, as more people start using them, there will be increased interest in finding weaknesses", he said.
Harley adds that, whilst botnets will continue to be a major problem, more people will realise that smaller low-profile botnets will pose as big a threat as the big named ones, which are monitored closely by security researchers.
This could, he predicts, result in the botnets being abandoned by their creators.
The ESET senior research fellow also makes the prediction that malware will continue to infect through the usual channels by tricking the victim into clicking on something ugly.
"However, there will be larger 'malvertising' campaigns, where individuals will click on advertisements from fake companies. It is also to be expected that unpleasant self-launching surprises like the LNK vulnerability, which was spread via networks shares and file-synchronisation systems, will also hit our radar from time to time, possibly long after the bad guys have discovered them", he explained.
So what about the Stuxnet worm?
Harley says that, whilst it is not quite the superbug that is sometimes suggested, the malware is pretty complex, since it takes a range of expertise, resources and sheer man-hours to pull off something so sophisticated.
"It's unlikely that the entire black hat community will unite in tiger teams to attack hard targets when there's lower-hanging fruit around. However, we've already seen a wide range of malware families 'borrow' vulnerabilities from Stuxnet", he said.
"These don't have the ambition and innovation or the sophistication of Stuxnet or ZeuS – this is just the bad guys adding an approach that seems to work for other attackers. The next big attack will probably be significantly different to Stuxnet, but it will come", he added.
One of the most interesting predictions that Harley makes is that there will be ongoing debate over anti-malware testing.
He argues that it's increasingly accepted that dynamic testing is a better representation than static testing of the current threat landscape as it affects AV users in real life, and the jury is still out on the ways in which to more effectively implement this form of testing.
"Testers and researchers within the anti-malware testing standards organisation will continue to play a prominent part in attempting to establish appropriate guidance, but some controversy is inevitable", he noted.