Firstly, Governments are reluctant to adopt Internet voting as long as it is seen as less secure than in-person voting. The argument is that as long as it is not fully secure it can do more harm than good. In other words, even if it is more convenient, easier and more accessible to vote on the Internet, but not fully secure, the entire legitimacy of the election process can be jeopardized.
 

Secondly, hacking, viruses and denial of service attacks can sabotage an entire election. Elections on the national, or state/provincial level are of course especially vulnerable because the stakes are higher. Thus they are more interesting targets for hackers. In addition, as elections are scheduled for a specific day, the window of opportunity is small and this also facilitates hacking (Cranor, 2001)

A third problem is voter identification. With remote Internet voting in use it is no longer possible to have polling officials verifying a voter’s identification. Internet voting must not make it possible for individuals to cast other than their own votes. Suggestions to overcome the identification problem include digital signatures and fingerprint or retinal scanning (California, 2000). However, even if the identification problem can be solved, the issue of complete ballot secrecy and reliable identification create a dilemma. How can an Internet voting system ensure both correct voter identification and ballot secrecy? In other words, the government needs to know who voted but not what he or she voted for (European Commission, 2001).

The final security concerns are vote tallying and vote tampering. The election software used must ensure the highest of accuracy while tallying the votes. This is crucial since Internet voting removes physical ballots and makes re-counts impossible. Tied to this is also the issue of liability. If the system fails, who is responsible, the software company that developed the technology, the consulting firm or the government? Moreover, voters also need to know that once they have voted nothing will happen to their vote (Cranor, 2001)

Internet voting may also open up the door for vote selling. This is of course more likely in countries where elections involve a lot of money, for example in the United States. Voter coercion or pressure must also be taken into account. Once polling stations are removed the voter is no longer as protected from the external environment. In other words, it becomes much easier to force people to vote a certain way. This does not always have to involve physical violence, but can be quiet subtle as in the form of pressure at work, or that a whole family votes ‘together’ (Olsson, itkommissionen, 2001).