The recent impact of generative artificial intelligence has led to businesses assessing whether to incorporate these technologies into their processes. However, how are artificial intelligence technologies impacting UK employees? Are they an opportunity or a threat? And how can generative AI help businesses?
In this article
Even before this launch, there were high expectations for this technology. According to Gartner, Venture capital firms have invested over $1.7 billion in generative AI solutions over the last three years. By 2027, Gartner predicts that nearly 15% of new applications will be generated by AI without human intervention. Nonetheless, it's to be expected that people might have concerns or seek more regulation in order to safeguard jobs.
However, as with many young technologies, organisations and employees must be aware of the opportunities, impacts, and challenges of deploying these tools and how to use them responsibly.
What is generative AI (GAI)?
GAI is a type of AI that can generate new and original content, such as images, videos, code or text. Like other AI tools, it can use deep-learning techniques and neural networks to analyse and learn from large datasets to generate content that resembles human creations. Since the launch of ChatGPT, Google and Meta have also accelerated the development of competing products, and by March 2023, Google had launched Bard in the UK as a direct competitor to OpenAI.
Capterra surveyed 496 employees who use GAI at least a few times per month for work to learn about the adoption, use and perception of generative AI in UK Companies. The full methodology is at the bottom of this article.
The vast majority of companies give some degree of importance to generative AI
Generative AI has the potential to transform the workplace. A recent report from Goldman Sachs stating that AI could increase global GDP by 7% also says that up to 300 million full-time jobs could be replaced due to AI.
Understandably, our study shows that most respondents (98%) say their company gives some importance to generative AI.
GAI can be applied to numerous employee tasks across various industries. This can include screenwriting, coding and content-driven jobs, but also different specialised sectors including healthcare, translation, design, finance, manufacturing, marketing, or sales, to name a few.
When we asked our survey respondents what they used generative AI for, the three most frequent uses were:
- Text editing (41%)
- Text creation (40%)
- Analytics and reporting (40%).
How to use GAI for text creation and analytics
Businesses can leverage GAI with AI writing assistant tools to help them generate content. These tools include features such as plagiarism detection and text and word suggestions to ensure that content is legible and grammatically correct. Many also directly feature automatic text generation based on minimal user input.
Generative AI can also assist in analytics and reporting by helping bridge gaps when analysing large data sets. GAI features in predictive analytics can create and test multiple hypotheses based on data sources, generating business insights and reports and updating insights in real-time, helping businesses make better-informed decisions.
9 out of 10 employees inform their company about generative AI use
Given the potential of generative AI and its diversity when it comes to applying it, our surveyed employees were asked if, apart from them, there is anyone else in their company who actively uses GAI. 61% of respondents answered that a significant portion of their company was actively using generative AI at work. Another 36% responded that some people were using the technology in their company, but that there was also a significant proportion that was not.
Furthermore, amid any suspicion that workers may be using these tools without their company’s knowledge, 93% of respondents said they had informed their company that they were using generative AI tools. This shows that this tool is generally accepted in the workplace.
How do businesses measure and control the quality of generative AI outputs?
While generative AI can efficiently create content from well-crafted prompts, it is not error-free. For instance, chatbots might generate false information and citations as if they were facts. These have been called AI hallucinations. Additionally, GAI can generate texts by pulling from multiple sources, but this can sometimes lead to the production of content that is the same as, or very similar, to content elsewhere on the web. To avoid publishing duplicate content, which Google advises against, businesses can use plagiarism checkers to ensure that their content is unique.
Ensuring that content is of high quality is paramount for businesses. Many companies regularly deploy quality checks of their products, and this should also be the case with generative AI content. Regular evaluations can also identify areas for improvement. Businesses should always bear in mind some of the drawbacks of GAI, including a lack of transparency, fabricated answers, deep fakes, bias, intellectual property violations or a failure to comply with GDPR.
Consequently, here are some methods to control the quality of the results:
- Collect employee feedback using surveys: 52% of respondents who openly use AI at work provide feedback and evaluations of the results of generative AI. Businesses can collect this staff feedback through surveys to track errors or recurring issues.
- Establish key performance indicators (KPIs)to monitor outputs: KPIs, such as engagement metrics, organic traffic or SEO ranking, can help businesses measure the performance of their AI-generated projects. These indicators can be presented in dashboards to measure performance in relation to targets and objectives. 47% of our survey group who openly used GAI used these methods to control the quality of their outputs.
- Compare human outputs with AI outputs: Businesses can determine which is more effective for specific tasks by comparing both types of outputs. 43% of employees who informed their companies about the use of GAI said that their company deployed this method to control the quality of their results.
- Select a dedicated team to evaluate the results: By choosing a team that is qualified and trained in artificial intelligence, businesses can leverage their expertise to accurately evaluate the results of their AI outputs. 37% of employees who communicated their use of GAI said that this is how their company checks the quality of their AI outputs.
- Hire an outside consulting firm to evaluate results: External consulting firms can provide AI experts with experience developing, implementing, and monitoring AI projects. By delegating these tasks, companies can save time and focus on other objectives. 15% of survey respondents who informed their managers of their use of GAI worked in companies where AI outputs were measured this way.
What are the benefits of generative AI?
What is it about generative AI tools that make it so appealing for businesses? There are several benefits for employees and organisations, but what aspects of GAI are the most effective? We asked our UK employees who used GAI to evaluate some of the benefits of using these tools.
Generative AI can offer innovative and creative outputs that save time and costs in research and development. By saving resources, businesses can gain a competitive edge and improve performance. However, caution is vital to verify content originality and ensure standards are met.
GAI is effective for analysis, creativity and productivity
GAI can drive employee efficiency in different ways. We asked our survey respondents to detail the three most effective ways they felt this technology has impacted their efficiency. The top three answers related to data analysis, creativity, and productivity:
Here are some ways GAI help businesses manage and get insights from their data.
- Data preprocessing: GAI can automate some tasks, like cleaning and structuring raw data. This saves time and effort in manual data preparation
- Insight delivery: GAI models can spot trends, correlations and anomalies in data, providing insights that can help in decision-making
- Summary generation: AI systems can generate summaries and highlight key findings from complex data
- Data visualisation: GAI can produce and interpret visual data representations, such as graphs and charts based on data
Exercise caution when inputting proprietary data into GAI systems
While GAI tools work from user prompts, there is a risk that these prompts can be reviewed by the GAI tool developers and used to train future models. If these prompts contain information that is already public, there may be little risk. However, if they include sensitive, proprietary, or confidential information, there may be the possibility of putting compliance obligations and intellectual property protections at risk. This has led to businesses like Amazon restricting the use of tools like ChatGPT, after it discovered that the AI technology was mimicking internal Amazon data in its results.
96% of respondents believe that generative AI increases their productivity
AI tools can enhance productivity through task automation, idea generation, content creation, and personalised assistance. Employees can leverage these tools to work more efficiently and achieve better results. We asked our respondents how AI influenced their productivity.
According to our data, 96% of surveyed employees said that generative AI increased their productivity to some degree.
This wasn’t the only way generative AI had transformed our respondent’s jobs. When asked to describe how their jobs had evolved due to GAI, we saw that:
- 46% of respondents had more time to focus on higher-value tasks
- 33% could interpret data quicker and more efficiently
- 16% could perform a wider array of tasks with these tools
- 4% were more confident facing new projects.
Only 1% of respondents thought that their job was not changing due to generative AI tools.
How is the use of generative AI being regulated in the UK?
The rise of generative AI has sparked multiple debates about legislation and guidelines. While the European Parliament swiftly launches one of the world’s first laws governing AI, the US is still lagging. Meanwhile, the UK Government recently shifted from a “pro-innovation” and “context-specific” approach that did not aim to create a new AI regulator, to a new stance where it seeks to play a leading role in global AI guidelines.
Some of the objectives of these regulations include addressing fake content, ensuring human supervision, and respecting security, privacy, transparency and energy requisites too.
Unlike the draft EU's AI Act, the UK Government does not aim to create a new AI regulator. The UK Government recognises this would cause complexity, confusion, and undermine the mandate of existing regulators. Instead, the UK Government plans to support existing regulators to apply the principles using the powers and resources available to them.
Regarding UK employees, how is the use of generative AI being regulated? According to our study, 72% of employees who informed their companies about the use of AI tools said that regulations and guidelines had already been implemented. Another 23% said they were planning to do so.
97% of respondents want some sort of guidelines for GAI
To delve further into regulation perspectives, we asked all respondents if guidelines for GAI usage at work were necessary. 68% favoured strict guidelines, and 29% agreed that some guidelines were required.
Guidelines can help employees understand best practices, and compliance with laws and regulations. Also, employees can undergo training to learn the appropriate use of the tools.
How to create GAI guidelines
As GAI becomes increasingly integrated into workplace practices, organisations should establish comprehensive guidelines to harness the potential of these tools and ensure responsible and ethical usage. Here are some tips for creating guidelines.
- Determine use cases and restrictions: clearly outline what employees can and cannot use these tools for. Establish any restrictions on generating sensitive or inappropriate content
- Outline input boundaries: to protect sensitive information and intellectual property, articulate what types of data employees are not allowed to input into GAI tools. This might include proprietary company data, confidential customer information, or any other classified material.
- Use tools to help create a clear policy: to streamline the process of developing GAI guidelines, leverage tools like policy management software. These platforms can help develop a structured framework for drafting, reviewing and approving policies.
- Share guidelines with the company: once the guidelines are finalised, share them widely within the organisation. Use internal communication channels, such as email, intranet or messaging platforms to encourage employees to review and familiarise themselves with them.
- Foster accountability: establish mechanisms for receiving feedback and addressing employee concerns related to GAI systems. Provide channels for end users and those impacted to raise concerns and ensure that the company is responsive to feedback.
Companies that adopt generative AI in the workplace can leverage these tools to bolster employee performance. However, care must be taken to ensure that employees are trained to use these tools ethically and that guidelines and regulations are set to ensure that content is original, and compliance is respected.
In the second part of this series on generative AI tools, we will evaluate the use of the AI chatbot ChatGPT and explore employee concerns about its usage in the workplace.
To collect this data, Capterra interviewed 496 UK employees online in June 2023. The candidates had to fulfil the following criteria:
- UK resident
- Between the age of 18 and 65
- Employed full or part-time
- Uses a computer/laptop to perform daily tasks at work
- Uses generative AI tools for their work
- Must have understood and correctly identified what generative AI is after being shown a definition
Respondents were provided with the following definition:
Generative AI (GAI) refers to a type of artificial intelligence that is capable of generating new, original content such as images, videos, music, code, or text. It typically uses deep learning techniques and neural networks to analyse and learn from large datasets and uses this information to generate content that resembles human creations. Some examples of generative AI tools are ChatGPT, Bard, and DALL-E.