AI Regulation: The UK Waits, But You Don't Have To

The UK government last week rowed back on its proposed AI '𝗼𝗽𝘁-𝗼𝘂𝘁'.

Instead it has pledged to create an economic impact assessment and to publish reports on key related issues:

1️⃣ 𝘁𝗿𝗮𝗻𝘀𝗽𝗮𝗿𝗲𝗻𝗰𝘆

2️⃣ 𝗹𝗶𝗰𝗲𝗻𝘀𝗶𝗻𝗴 and

3️⃣ 𝗮𝗰𝗰𝗲𝘀𝘀 𝘁𝗼 𝗱𝗮𝘁𝗮

before taking any further action to changes to the current laws.

With reports that this could delay the process to 2029, an abundantly unworkable timeline given the global nature and pace of development, uncertainty in the law continues to prevail.

For the time being, it is likely the government will lean towards 𝘀𝗲𝗹𝗳-𝗿𝗲𝗴𝘂𝗹𝗮𝘁𝗶𝗼𝗻 by empowering its existing market regulators to oversee and enforce 'purpose' and 'outcome' driven market standards.

𝗛𝗼𝘄 𝘁𝗼 𝘀𝘁𝗮𝘆 𝗮𝗵𝗲𝗮𝗱?

If you're someone who works in the intersection between AI and the creative industries or simply a UK business using AI tools, the key issues ☝️ are big clues on how to comply with market standards.

In practice, this means:

🛡️𝗦𝗮𝗻𝗱𝗯𝗼𝘅𝗲𝗱 𝗔𝗜 𝘁𝗼𝗼𝗹𝘀: that ring fence inputs, data, and proprietary information etc = 𝘢𝘤𝘤𝘦𝘴𝘴 𝘵𝘰 𝘥𝘢𝘵𝘢

📜𝗔𝗜 𝗲𝘁𝗵𝗶𝗰𝘀 𝗽𝗼𝗹𝗶𝗰𝘆: to ensure ethical and legal adherence - think of a privacy policy but for AI = 𝘵𝘳𝘢𝘯𝘴𝘱𝘢𝘳𝘦𝘯𝘤𝘺

👁️‍🗨️𝗔𝗜 𝘁𝗿𝗮𝗰𝗸𝗲𝗿: document your AI tools - so if you're asked, you have it on you, an Excel spreadsheet will do = 𝘵𝘳𝘢𝘯𝘴𝘱𝘢𝘳𝘦𝘯𝘤𝘺

🤝𝗟𝗶𝗰𝗲𝗻𝘀𝗶𝗻𝗴: using ethically licensed tools is better than an indiscriminate scrapers = 𝘭𝘪𝘤𝘦𝘯𝘴𝘪𝘯𝘨

Think of the above as an 𝗔𝗜 𝗶𝗻𝘀𝘂𝗿𝗮𝗻𝗰𝗲 𝗽𝗼𝗹𝗶𝗰𝘆.

Most of the market will show up too late. You won't.