Work on improving security in the Internet of Things (IoT) continues apace! The UK government has reached another milestone in its mission to make the country one of the most secure places to do business and to live in, with the release of proposals for regulating the cyber security of smart products. They are well worth a read and provide a good steer as to what the future of insecure connected products will look like when we collectively say ‘Enough is enough’.
This Call for Views invites feedback until early September 2020 on a range of options as the government moves towards legislation based around the top 3 items in the UK’s Code of Practice for IoT Security:
1) To eliminate the problem of default passwords.
2) To ensure that companies in the IoT space have a way for security researchers to be able to contact them to report vulnerabilities in products.
3) To be transparent to consumers about how long software updates will be available.
These are anchored in the recently approved European standard for IoT security, ETSI EN 303 645 which has the support of industry and governments across the world, marking a significant harmonisation of views on how the problem should be approached.
The Call for Views outlines the aims of the government – to achieve an outcome where there are no products available on the UK market that are non-compliant with the above. In simple terms – you shouldn’t be able to buy a product that has not been designed securely.
This is of course just the start. The items above are fundamental, but there many different types of security that should be built into products, it’s just that some manufacturers of products and services choose not to do that. You wouldn’t allow a food manufacturer to supply to shops if they hadn’t taken basic sanitation measures so why should that be allowed in the smart product space?
The scope of products included is broader than IoT products and covers the scope of nearly all the connected products you could find in a home, including laptops and mobile phones. As PCs and mobile phones have been under attack for many years now, the product security in those industries is significantly mature and it really shouldn’t be an issue for those companies to conform to the basics because they’re already doing them.
The core scope is the connected products that everyone has concerns about – children’s toys, cameras, appliances such as fridges or washing machines, safety-relevant products such as connected door locks and so on as well as IoT ‘hubs’.
One area that has been a significant concern for many years is home routers. These rarely get updated and often stay in place in homes for many years without being touched. If they’re compromised, they can create a big issue to users because they’re the point of entry to the home and everything else that is connected, but equally, compromised routers and other equipment at scale can create harm to others across the world by being part of other types of attack.
The proposed scope also covers home workers by including things like printers and office equipment that you might find in both a home or office. This is particularly relevant as businesses have shifted their workforces to home during the Covid-19 crisis.
Things that are out-of-scope are because there is existing or forthcoming regulation in those domains – for example, smart Electric Vehicle (EV) Chargers, Smart Meters and medical devices.
The work laid out in the proposals sets out the obligations on Producers and Distributers, formalising the language that has been used thus far such that it forms the basis of a legislative and regulatory framework governing people who make products but also those that sell them into the UK. It also means that there must be a way to test and declare compliance of these products. This comes at a good time as the EU Cyber Security Act will also require such action to take place across lots of different types of networked products. The proposals also lay out when they expect companies to be compliant – it is proposed that everything must be in place by 9 months following Royal Assent of legislation. The implication is that companies have had long enough and enough warnings that these practices are simply not acceptable.
The list of proposed enforcement actions aligns with other existing ways of removing products from the market – i.e. issuing compliance notices, through to enforcement with real teeth: it is proposed that order breaches are contempt of court which carries a maximum penalty of a fine and two years’ imprisonment. Forfeiture and destruction of products are also on the table as well as financial penalties – the fine amounts are to be determined but a note states that other regulations consider fines of up to 4% of annual worldwide turnover (a clear reference to the EU data protection regulation GDPR). This alone shows that the intent is for the regulation to have real teeth and that the government means business. The ‘Avengers’ team of superheroes working on this project at DCMS and NCSC have done a fantastic job once again, supported by lots of other government departments. Especially now as well – ‘Quiet Batpeople’ is certainly not the right term, but these individuals have all also been volunteering to deal with various aspects of the Covid-19 response, so to deliver this work as well is a huge achievement!
Mapping the Global Direction and Understanding of IoT Security
Understanding where everyone stands on this from a technical perspective is a tough job. I am lucky to have a fantastic team who have been working on doing just that. We have continually been monitoring the progress of IoT security recommendations and standardisation and will continue to do so. Our work can be seen at https://iotsecuritymapping.uk. We recently added recommendations from Australia, Singapore, California’s new law on connected device security and the US National Institute of Standards and Technology (NIST)’s Device Cybersecurity Capability Core Baseline. There are more documents being mapped soon and we’re tracking work from Brazil, to India, to proposed legislation in the US State of Oregon.
We have noticed that there is defragmentation of ideas and recommendations happening across the world as there is a greater collective understanding of the problem domain and how to solve it. The mappings that we have recently created show strong alignment against the top 3 items listed above. We have also observed that whilst some countries are slightly less mature than the UK in tackling the issue, they can benefit from the international standardisation that has taken place and are starting to adopt and endorse this already. Truly we can adopt a global stance that it is unacceptable to provide connected products without even considering the basics of product security.
Here’s some more background material if you’re interested in further reading:
- Legislating for Security in Consumer IoT
- Mapping New IoT Security Recommendations
- ETSI publishes European Standard on Consumer IoT Security
- Stepping up action on IoT insecurity – new laws and regulation
- Investigating the State of Vulnerability Disclosure in Consumer IoT Products
- Security change for good in the Internet of Things
- Consumers should be able to reject IoT products as not secure with these simple checks
- Mapping IoT Security and Privacy Recommendations and Guidance
- Discussing the UK Government’s Code of Practice for Consumer IoT Security and the Future (audio)
- Code of Practice for Security in Consumer IoT Products and Services – David Rogers at 44CON 2018 (video presentation)
- Government Reports, IoT Security, Mirai and Regulation
- A Code of Practice for Security in Consumer IoT Products and Services
- How the UK’s Code of Practice on IoT security would have prevented Mirai