The flaws were published on a website, accompanied by a white paper from which technical details were redacted, in order, CTS Labs said, to prevent exploitation based on what it had decided to publicise. From archived information, it appears that the CTS Labs domain was registered in June last year.
AMD was given 24 hours to respond though standard practice in the security community for many years has been to give a company, whose products are found to have bugs, 90 days to respond.
In most cases, disclosure after the 90-day period is a exercise which is co-ordinated between the bug finder and the company whose product is vulnerable.
The disclosure of the bugs comes a couple of months after flaws in Intel processors, known as Meltdown and Spectre, were made public. This disclosure, however came after a long period, as the initial information was provided to Intel by Google in June 2017.
The bugs themselves fall into four categories (names given by CTS Labs): Ryzenfall, Chimera, Fallen and Masterkey.
CTS Labs said that the Ryzen chipset, a new line from AMD, was being shipped with exploitable backdoors, which had come about as a result of obtaining technology from ASMedia, an outsourcing partner.
Yeah tinfoil hat machines are running crazy right now, for now lets just keep at it that its a shady company seemingly wanting to throw shade on AMD. Majority of vulnerabilities seem irrelevant, good summary by @cynicalsecurity: https://t.co/3tVmNU4JUs— Yonathan Klijnsma (@ydklijnsma) 13 March 2018
It pointed out that ASMedia, a subsidiary of ASUSTek Computer, had been penalised by the US Federal Trade Commission for not taking security seriously.
Under the Ryzenfall category, CTS Labs claimed that malicious code could be used to take over the AMD Secure Processor; privileges of this processor could be used to write into protected memory areas; Windows Credential Guard could be bypassed and network credentials stolen; and Ryzenfall could be used along with Masterkey to install persistent malware on the Secure processor.
In the Fallen category, CTS Labs listed vulnerabilities that allowed reading and writing to protected memory areas; leveraging of these flaws to steal network credentials protected by Windows Credential Guard; and bypassing BIOS flashing protections implemented in SMM.
Under the category Masterkey, CTS Labs claimed that there were a number of flaws in the firmware of the secure processor that would allow attackers to gain access to this processor; stealthy and persistent malware could reside in this area; AMD's firmware-base security features like Secure Encrypted Virtualisation and Firmware Trusted Platform Module could be tampered with; network credential theft was possible, and hardware could be physically damaged and bricked.
In the Chimera category, CTS Labs said it had found two sets of backdoors, one in firmware and one in hardware, both allowing malicious code to be injected into the Ryzen chipset; the chipset's middleman position could be leveraged for attacks; chipset-based malware could evade endpoint security solutions; and malware on the chipset could use direct memory access to attack the operating system.
Someone found their greenscreening backdrops, why would you even want your real office there? pic.twitter.com/HuT8qsU5yL— Yonathan Klijnsma (@ydklijnsma) 13 March 2018
All the vulnerabilities were revealed to the security firm Trail of Bits last week. Its founder Dan Guido told the Motherboard website that each of them was exploitable and worked as described. Guido was paid US$16,000 for his work as a contractor, a fact he disclosed in a tweet.
I initially responded to their request out of curiosity -- "Hey, do you want to see our new processor bugs before we release them?" "hell yes I do" -- but after their asks continued to grow billed them our week rate for the work.— Dan Guido (@dguido) March 13, 2018
British security researcher Kevin Beaumont said in an initial technical analysis:
- All of the bugs require administrator (or root) access to exploit. This is a significant mitigation.
- All of the bugs require the ability to execute code. This is a significant mitigation.
- No proof of concept code has been provided.
- No technical information has been published.
- Nothing is in the wild for this.
- It could not lead to a global cyber attack like WannaCry, as it does not provide code execution.
iTWire has sought clarification from CTS Labs on various aspects of the disclosures, including matters surrounding them which have been raised by various security researchers.
AMD said in a public statement on its website: "We have just received a report from a company called CTS Labs claiming there are potential security vulnerabilities related to certain of our processors.
"We are actively investigating and analysing its findings. This company was previously unknown to AMD and we find it unusual for a security firm to publish its research to the press without providing a reasonable amount of time for the company to investigate and address its findings.
"At AMD, security is a top priority and we are continually working to ensure the safety of our users as potential new risks arise. We will update this blog as news develops."