In our June edition of Cyber Security Partner’s Newsletter this month, we go over Regulation and AI, which highlights the importance of advocating responsible AI design, development and use, and legislation that will regulate the use and outcomes of AI. Then we discuss how to keep your bank account and children safe whilst using apps on your devices.
David Bishop
Regulation and AI
Despite the rapid development in areas of Artificial Intelligence (AI), and growing interest around its seemingly unchecked deployment across industry sectors, there is still a gap where well-defined regulations would be welcomed. Addressing legal and ethical issues is of great importance in areas of privacy and surveillance, bias or discrimination, and the role of humans in processing activities where decision making could be critical to the health of citizens and the economy.
In terms of data protection, AI can bring benefits through automation and efficiencies, but also introduces a new source of inaccuracy and an unwelcomed driver of data breaches. Amongst the posturing within the cyberspace race, there appears to be an unwritten agreement that the transparency of algorithms is paramount to the rights and freedoms of individuals, and fundamental to the development of cybersecurity capabilities to run alongside emerging vulnerabilities.
In April 2021 a proposal was tabled for a regulation of the European parliament to harmonise rules on AI (Artificial Intelligence Act) and amending certain union legislative acts. In May 2023 the European Parliament’s Internal Market and Civil Liberties committees adopted the AI Act with a broad majority, which now progresses to a further vote in mid-June. Highlights of the Act include:
- A definition of AI which is aligned with that of the OECD (Organisation for Economic Co-Operation and Development).
- A list of prohibited practices has been expanded upon, stricter rules are in place for foundation models and generative AI.
- An extra layer of control for high-risk categorisation, more substantial obligations for high-risk systems and the establishment of centralised enforcement mechanisms.
Meanwhile, in the UK there is a similar looking Government consultation and policy paper for a pro innovation approach to AI regulation. On the face of it, this states an overriding intention to maximise AI, however the protections can also be found just inside the front cover. It promotes the need for a proportionate, context-specific approach to regulation, with impact assessments as a tool for managing liabilities throughout the AI life-cycle, and regulatory enforcement balanced against technical tradeoffs.
It comes as no surprise that the consultation advocates responsible AI design, development and use, and legislation that will regulate the use and outcomes of AI, rather than the technology itself. An increase in public trust will be needed if the government are to harness the strengths of AI, while ensuring its application is proportionate, adaptable and content-sensitive.
Lynne Mowles
Children using apps
I recently read a news story about a 10-year-old who had managed to make app purchases on a shared family device. It made me think about children that I know who use smartphones and electrical devices and how much they understand.
I spoke with a 10-year-old in our family and asked them about the particular app that the child had been using. The first thing he said to me was that before giving a child access to a device that would be able to communicate with others and make purchases, it would be to teach the child about online safety. Luckily, his device is one that we have set up ourselves and one that is secure. Whilst he does not have family sharing with a parental Apple ID, he has his own which we have the password for and there is no payment card linked to his account.
I asked him about the particular gaming app he uses, we opened the app together and I asked him to show me what it looks like when he tried to make a purchase. The first thing that it did was to flag up the following warning ‘This purchase involves the exchange of real money’. It then asks for Apple ID and Password, if there had been a payment card linked to his Apple ID, the payment would have gone through.
I then went online with him, and we did some research as to how the purchases could have been prevented. In this case it was a shared family iPad that was used to make the purchases. There are various ways that this could have been prevented:
Option 1 – If the device belongs to a child or a child will be using the device, you should prevent other users from changing the settings. You can do this by tapping ‘Use Screen Time Passcode’ to create a passcode meaning only you can change settings etc. as your Apple ID and password will then be required. It is important to keep your Apple ID and Password safe and not share it with others.
Option 2 – Use ‘Ask to Buy’ which screens and requires approval for purchase requests. To use this option, you would need to set up Family Sharing. This gives parents and guardians control over app purchase requests.
The key thing that I took from the conversation that I had, was that not only should you teach your children how to use devices but also to ensure they know how the games they are playing work. You should disable pop ups where you can and encourage them not to click on advert links as this can take them to other sites that might have adult content such as gambling and dating sites. You should make sure you know what the apps do that they want to use by reviewing them before you allow them to download them.
There are a lot of teaching resources out there to help parents and guardians to help them make safe online decisions.
Teach your child about safety online and explain to them that the app purchases are just the same as going to the shop by taking actual money from the bank account that is linked to the device.