Generative AI, exemplified by tools like ChatGPT and Midjourney, has swiftly emerged from obscurity to seize global attention. Ofcom’s experts, including Benedict Dellot, Anna-Sophie Harling, and Jessica Rose Smith, unravel the organization’s response to this transformative technology.
As generative AI, led by ChatGPT’s meteoric rise, disrupts industries weekly, Ofcom meticulously gauges its implications. This article explores the profound impact on communication sectors, from the fastest-growing app status to potential risks. Amid debates on positive change versus risks, Ofcom’s strategic initiatives and collaborative efforts with regulators shape a balanced approach to this evolving industry.
To gauge the rapid pace of developments in the generative AI realm, one only needs to observe the proliferation of new models each week and the substantial influx of capital into AI startups. The preeminent model, ChatGPT, garnered over 100 million users within two months, marking it the fastest-growing consumer internet app. Comparatively, it outpaced TikTok, which took nine months to reach the 100 million user milestone, and Instagram, which achieved it in over two years.
Table of Contents
The Implications for the Communications Sector
Transformation in the Communications Sector Regardless of whether one views generative AI as a force for positive change or as a source of potential risks, consensus among experts suggests its substantial impact on the future of our economy and society.
This impact is particularly pronounced in the communications industries, spanning telecom security, broadcast content, online safety, and spectrum management, where generative AI has the potential to disrupt conventional service delivery, business models, and consumer behaviour.
Several of these disruptions carry positive prospects. In TV content production, generative AI models enhance producers’ capabilities to craft visually compelling effects. In the realm of online safety, researchers explore the utility of generative AI for generating synthetic training data, thereby improving the precision of safety technologies. By identifying abnormalities on a network, generative AI can flag potential malicious activity, bolstering data and online asset security.
However, the adoption of generative AI also introduces risks. Such tools could exploit voice clones generated for phone scams and impersonating loved ones. Fraudsters might leverage generative AI to create more sophisticated phishing content.
Furthermore, generative AI could pose risks to online service users, providing easier access to self-harm instructions or advice on smuggling illegal substances.
Concerns extend to the creation of ‘fake’ news and media through generative AI models, precipitating rapid online dissemination. This poses challenges for broadcast journalists tasked with authenticating content from online sources.
Moreover, there’s apprehension that these tools may inadvertently produce inaccurate or biased news content, potentially undermining efforts to cultivate a pluralistic online news ecosystem.
Ofcom’s Strategic Approach to Generative AI
Teams across Ofcom closely monitor the rapid evolution of generative AI. The technical, research, and policy units engage in thorough research to better comprehend the unique opportunities and risks surrounding the development and utilization of generative AI models within the communication sectors regulated by Ofcom.
A particular focus lies on understanding the measures developers and industry stakeholders are implementing to mitigate potential risks.
Key Initiatives:
- Collaborating with companies incorporating generative AI tools that may fall under the Online Safety Bill, seeking to understand their proactive safety assessments and the implementation of effective mitigations to safeguard users.
- Vigilantly observing the impact of emerging technologies, such as generative AI, augmented and virtual realities, on people’s media literacy.
- Providing information for regulated sectors on the implications of generative AI, ensuring clarity on responsibilities to customers. For example, recent advice was issued to UK broadcasters, clarifying how the Broadcasting Code applies to the use of synthetic media.
- Scrutinizing evidence regarding detection techniques distinguishing between real and AI-generated content. Exploring the role of transparency, exemplified by the Content Authenticity Initiative standard, in indicating whether content is developed by humans or generative AI models.
- Actively participating in international think-tank discussions on AI regulation and contributing to multilateral expert groups to shape best practices for ethical AI use in journalism.
- Continuing efforts to understand generative AI, including hosting a dedicated ‘tech week’ with external speakers to discuss technological advancements and risk-mitigation measures.
Collaborative Efforts
Ofcom aligns efforts with digital regulator partners through the Digital Regulation Cooperation Forum. This involves hosting internal and external discussions to share research on generative AI and explore opportunities for collaborative endeavors.
Next Steps
Various stakeholders in this industry are actively working to harness the benefits of generative AI while minimizing potential risks. As companies integrate generative AI models, an expectation exists for them to assess associated risks and potential threats, implementing transparent systems and processes to build confidence in risk mitigation. Ongoing engagement is encouraged by those involved in generative AI development and integration as we navigate these critical issues.
