The discussion about how law enforcement or government intelligence agencies might rapidly decode information someone else wants to keep secret is shifting. One commonly proposed approach, introducing what is called a ‘backdoor’ to the encryption algorithm itself, is now considered too risky.
The research community and technology industry appear to be in agreement that weakening the encryption that in part enables information security – even if for national security – is a bad idea.
What comes next? Surely intelligence agencies will still want information stored by criminals in encrypted forms. Without a backdoor, how will they get access to data that may help them solve or even prevent a crime?
The future of law enforcement and intelligence gathering efforts involving digital information is an emerging field, sometimes called ‘lawful hacking.’ Rather than employing a skeleton key that grants immediate access to encrypted information, government agents will have to find other technical ways – often involving malicious code – and other legal frameworks.
Decades of History
In the mid-1990s, the Clinton administration advanced the Clipper Chip. The chip, which ultimately was doomed by its technical shortcomings, was an attempt to ensure government access to encrypted communications. After the chip’s failure, a group of cryptographers formally studied various mechanisms that might allow a trusted third party to read encrypted data in emergencies. They concluded that each approach had significant security risks.
The cryptographers’ view was that introducing this new capability into an encryption system made an already complicated process even more complex. This increased complexity made it more likely that there would be an unintentional vulnerability hidden in the encryption protocol that malicious hackers could find, gaining access to the trusted third party’s emergency system or otherwise breaking the code. The hackers could then read secret messages for their own purposes – a huge risk.
While the Clipper Chip effort to use public processes to create weaknesses in cybersecurity failed, the National Security Agency had, in secret, worked to undermine certain popular encryption algorithms. In addition to direct attempts to break encryption with mathematical methods, an NSA project code-named Bullrun included efforts to influence or control international cryptography standards, and even to collaborate with private companies to ensure the NSA could decode their encryption.
This came to light when former NSA contractor Edward Snowden revealed a massive trove of files about the US government spying in 2013 and reignited the debate about what abilities and powers the government should have to read encrypted material.
Again a group of the world’s leading cryptographers studied the issue, and in 2015 came to the same conclusion: The risk of backdooring encryption to enable government access was too high. Doing so would weaken overall security too much to make up for any brief improvements in public safety or national security.
FBI Pushes Back
On Dec 2, 2015, Rizwan Farook and his wife, Tashfeen Malik, opened fire at a social services centre in San Bernardino, California. Inspired by foreign terrorist groups, they killed 14 people and wounded 22.
Before the attack, Farook had smashed two personal cellphones, rendering their data unrecoverable. He left untouched his work phone, an iPhone 5c issued by San Bernardino County. Investigators found the phone, but the FBI was unable to examine its data due to Apple’s encryption and security mechanisms.
To get around this, the US government used a law from the earliest days of the republic, the 1789 All Writs Act, to try to compel Apple to write software that would break the encryption and grant the FBI access. Apple refused, saying that doing so would weaken the security of every iPhone in the market, and a court showdown began.
The Apple-FBI case nicely encapsulates much of the debate around encryption: a horrible incident that everyone wants investigated, the government’s stated need for access to aid the investigation, strong encryption that prevents that access and a company unwilling to risk the broader security of its products by attacking its own software.
Faced with Apple’s refusal to comply and criticism from the technology and privacy industries, the FBI found another way. The bureau hired an outside firm that was able to exploit a vulnerability in the iPhone’s software and gain access. It wasn’t the first time the bureau had done such a thing.
As this all unfolded, and in the face of a wide range of significant opposition, a Bill to mandate backdoors was introduced and failed in the United States Congress.
Encryption backdoors remain largely viewed as weakening everyone’s protections all the time for the sake of some people’s protections on rare occasions. As a result, workarounds like the FBI found are likely to be the most common approach going forward.
Technologists and lawyers studying the issue have identified several key questions, but not their answers. These include:
· What kinds of vulnerabilities can law enforcement use to gain access, technologically, legally and ethically?
· Should they report those vulnerabilities to the software vendors for fixing, even if it means it is less likely that either police or hackers will be able use the weaknesses in the future?
· What do they need to tell a judge in order to get permission to hack a device?
· Can they hack devices outside of their jurisdiction, and what happens if they hack computers in other countries?
· Do they need to tell a defendant at trial how they hacked his or her device?
While some details depend on specific certain answers to these legal and technical questions, a lawful hacking approach offers a solution that appears to gain greater favour with experts than encryption backdoors. It’s time to turn from the now-ended debate about encryption backdoors and engage in this new discussion instead.