This point has come up a lot recently and although I do not intend to get into any great detail on the subject in public, I do need to highlight some issues which this single suggestion encompasses well. Some of the points I wish to raise are similar to what would be discussed between staff and the very highest echelons of Silk Road when we consider new ideas. I'm not aware of who first suggested it, but there is word spreading that multi-signature addresses are somehow the holy grail on online markets and that it would stop bugs, thefts and problems faced from law enforcements, but this severely overlooks a lot of the practical problems we face as an online market. Yes it is true that it is theoretically more secure than the current bitcoin system as it is backed by strong crypto - but in practice it does not convert so well given steps must be taken in addition to usual market practices. Right now, as an approximate market figure, PGP uptake is between 8% and 12% (unknown reason for fluctuations but US timezones have the lowest PGP uptake of all). PGP is a relatively straight forward step towards protecting your privacy and as far as the learning curve goes, it is certainly below that of learning to sign multi-signature transactions. So the evidence right now to suggest implementing the idea would be more beneficial than harmful is not true as we find preventing customers from having an easy to use interface will drive them to other markets, ones which are usually less secure and less trustworthy as we have seen with Sheep marketplace. Do not forget, Sheep was cast the eye of doubt when it first come about but because it was an easier to use interface, it took the lions share of the market above Black Market Reloaded which has a long and stable history. Another concern highlighted by my fellow staff is that to sign a transaction you have two options - local storage or remote services. Remote services would negate the point of the implementation so for this we shall disregard that consideration leaving us with local storage. If a person was to have their hardware seized, holding a bitcoin wallet will then tie you in to the block chain which cannot be concealed and is certainly one of the strongest pieces of evidence which could be presented against an individual. This risk can be significantly mitigated through the use of encryption layers but to the average user it is unlikely they would take such measures and so we are only actually leaving more evidence in less secure places. Multi-signature transactions will also mean the transaction must be known to all 3 parties and therefore the bitcoin trail faces increased exposure. Silk Road operates and protects users by making it impossible for one party to identify which deposit or withdrawals belongs to which buyer or vendor, something not possible with multi-signature transactions. It could be argued that mixing services may be employed to mitigate such risks but our research has shown only a fraction of users actually mix their bitcoins effectively before placing them into our system and so I have no confidence in such a setup as it provides further evidence for "honeypot vendors" should they ever be set up looking to arrest buyers, and it further prevents lazy vendors from being caught if they do not take the proper precautions. Most users do not mix their bitcoins, but an even smaller user share is those who mix their bitcoins without a single point of failure (ie use more than 1 company/service to mix them) and so if law enforcement were to gain leverage over a mixing service or continued their program of illegally tapping into such services (a credible threat) then this brings the risk of huge exposure for tens of thousands of users. I hope this has helped clarify at least some of the reasons the administration is reluctant to implement this suggestion in its current form. We do have some ideas to counter some of the above ourselves, but even with that it is difficult to see this as a step in the right direction just yet. We are all for making use of proven technology, but at the same time we must make it as accessible as possible to not drive away those who aren't comfortable with it just yet.