• AI is happening fast and even experts are experiencing an overwhelm of developments.
• AI could be better described as ‘Applied Statistics’.
• ChatGPT has taken AI into the mainstream because it’s delivered the first human experience of working with a machine.
• Mo Gawdat, Chief Business Officer at Google [X], believes by 2026 we’ll have created something as intelligent, or more intelligent than humans. By 2037 AI will have superintelligence.

How does ChatGPT learn?

• It trains itself on data sets (books, websites, articles) and on user inputs.
• User input is a red flag for cyber experts and lawyers.
• ChatGPT recently introduced the option to turn off the training function.
• To do this, open ChatGPT, click on your name in the bottom left, select Settings, select Data controls, toggle Chat History and Training to off.
• Is it now safe to input sensitive data in chatGPT? The answer is no. Don’t input personal or commercially sensitive data.

4 tips to maximies use of ChatGPT4:

1. Start your prompt with ‘Act as’, for example: ‘Act as a sales executive…’. Learnprompt.org has a library of prompts or the Prompt Perfect plug-in (for premium users) will guess what you’re really trying to find.
2. Brainstorming partner – useful to get a second view.
3. Don’t forget the ‘continue’ command.
4. Drop in a microphone option using a Google Chrome extension.

Top 5 plug-ins for premium ChatGPT users:

• Web pilot
• Prompt perfect
• Zapier
• World news
• Voxscript

5 practical steps that employers may wish to take:

• Communicate early with employees about AI use.
• Create an AI Use Policy/training and update other related policies.
• Stress the importance of double-checking ChatGPT suggestions for accuracy, potential for bias/descrimination.
• Be careful when using personal or commercially sensitive data.
• Always think of data protection.

An AI Use policy will:

• Create a consistent approach.
• Set standards and expectations.
• Provide a framework for employees.
• Help you realise the full potential of AI while minimising risk.

Training staff on data protection principles is important:

• The current legislation for data protection does not have ChatGPT in mind.
• It’s essential your policies are up to date and that these are communicated to all employees.
• Your Privacy policy should be reviewed and updated in line with changes in AI.
• Look into wider training if necessary.

Other areas where legal issues may arise when using AI:

• Copyright
• Right to be forgotten
• Right to correct inaccurate information
• Defamation

Useful AI resources:

• Littler Report on AI: a risk audit article pages 46-47 
• Podcasts: London Real – Mo Gawdat & Janet Adams on Zarqar
• Podcasts: The AI breakdown podcast; This Day in AI podcast

ChatGPT is already changing how we work. But without regulation, this powerful new tool can potentially be misused and abused in the workplace.

While recent plug-ins promise to enhance the performance and impact of ChatGPT even further, unregulated use has the potential to lead to litigation.

Spend 60 minutes in the company of industry experts and leave this webinar with:

Key takeaways to protect employees from misuse and abuse of this new tool.

Strategies to optimise ChatGPT’s potential in a safe and productive way.

Barry Phillips is Chairman of Legal Island a company that has been supporting HR professions in the UK and across the island of Ireland since 1998. Previous to that he worked as a barrister in London specialising in employment law before working as a legal adviser in both Belfast and Dublin.

Sarah Cochrane is a partner working in the Employment law team at Carson McDowell in Belfast. She provides practical employment advice to employers from a wide variety of industries throughout the UK. She is recognised as a stand out name by clients in the Legal 500 industry guide.