The security spotlight has been firmly thrust on the potential for - and technical feasibility of - supply chain attacks after Bloomberg claimed tiny “malicious chips” were added to server motherboards during manufacturing to alter their normal operation.
Bloomberg’s Businessweek last night reported more than 30 US companies were targeted in the attack where “a tiny microchip, not much bigger than a grain of rice, that wasn’t part of the boards’ original design” was inserted by contract manufacturers.
The boards were allegedly supplied to the US-based server motherboard company Supermicro, and ended up in enterprise customers’ servers, Bloomberg reported.
Supermicro - and a selection of current and former large customers - have all denied ever finding “malicious chips” that had been incorporated into the motherboard during manufacture.
“Supermicro has never found any malicious chips, nor been informed by any customer that such chips have been found,” the company said in a statement.
Apple similarly said that it “has never found malicious chips, ‘hardware manipulations’ or vulnerabilities purposely planted in any server.”
“As a matter of practice, before servers are put into production at Apple they are inspected for security vulnerabilities and we update all firmware and software with the latest protections,” Apple said.
“We did not uncover any unusual vulnerabilities in the servers we purchased from Supermicro when we updated the firmware and software according to our standard procedures.”
Likewise, Amazon said it had “never found modified hardware or malicious chips in servers in any of our data centres.”
Even if the so-called “malicious chips” had not been identified and “found” to date, there will likely be extensive efforts made over coming weeks to make sure they don’t - or didn’t - exist.
But there is also a wider question now of whether they could exist, and if so, how they would technically function.
Hardware security pentester Joe Fitz is one of the first domain experts to attempt to unpack the feasibility of the reported Supermicro compromise.
Fitz applied “a technical and feasibility lens” to the claims and said he was “confident there’s some truth to the story.”
He said that while hardware implants could be performed during manufacture, they were “rare”, in part because there are “plenty of software vectors” available that are less complex and that did not leave behind a physical item that could be discovered - but could achieve a similar result.
It has since emerged that the same attackers behind the alleged hardware compromise may also have used software-based vectors.
Fitz was also unclear about how the malicious chip communicated back to the nation state that allegedly planted it - a topic other security researchers have also speculated on.
“Every board has it, but we probably only care about one targeted customer of the board. This is where it gets complicated. If 10 million backdoored motherboards all ping the same home server, everyone will notice. I don’t have a solution here,” he said.
Nicholas Weaver, a senior staff researcher focusing on computer security at the International Computer Science Institute, published his own detailed take on the alleged attack.
“The attack described in the article is actually plausible,” Weaver said.
“I expect we will see independent confirmation of this attack within a few weeks.”
Weaver said that as “modern circuit boards are filled with small support chips ... the backdoor chip would appear to be just another faceless component to all but the most detailed examination.”
He theorised that the attack could be aided by the “serial EEPROM chip or a serial FLASH chip, which is used to store program and other instructions used during the startup process”.
White hat lcamtuf tweeted caution at the Bloomberg report, though he noted that on a broader level, it is “still prudent to worry about hardware supply chains.”
Other infosec researchers reserved judgment pending the expected release of more data that would either confirm or deny the existence of the “malicious chips”.
