PDF logo

Applications of Artificial Intelligence (AI) to Publishing & Training Services

TOPICS: AI and Publishing | ChatGPT | Software Updates

AI is a hot topic at present, particularly in the publishing sector. There is increasing concern about the ability of self-trained large language models (LLMs and GPTs) to generate 'realistic' student essays, journalist articles, books and academic papers. But can such systems provide reliable and accurate tools that might benefit publishers, training organizations and educational institutes?

Back in 2020 The UK Publishers' Organisation produced a report on the role of AI in Publishing as they saw it at the time. In it they identified a number of application areas for the technology, most of which were ill-understood and barely developed by the industry at the time. The applications they identified included:

  • Content extraction and text summarisation
  • |
  • Content creation
  • |
  • Identifying market trends
  • |
  • Language checks
  • |
  • Plagiarism checks
  • |
  • Copyright infringement
  • |
  • Recommendation engines
  • |
  • Demand forecasting
  • |
  • Stock management
  • |
  • Automatic invoicing
  • |
  • Chatbots
Of these application areas, most are now developing into significant products and services - from relatively 'simple' products (such as Adobe's content structure Extract API), to sophisticated 'bot' services such as the latest versions of ChatGPT, Quillbot and the MidJourney image creation service (see screenshots below), to complete AI-generated books including AI-generated graphic images. And not surprisingly, many businesses claim to provide AI-driven products and services in order to gain market traction, even if in reality they have little relevance or direct bearing to what they deliver.


There are currently over 200 books sold on Amazon that list ChatGPT as author or co-author and and unknown number of other publications whose full or partial authorship has been AI-based (an example ChatGBT response on AI in Publishing is provided below - repeating a request will often produce a new response, which may not be entirely in line with the previous one). And, not surprisingly, a range of additional products and services have sprung up - amongst these are training- and education-focused applications, such as McGraw-Hill's SmartBooks Adaptive Learning System and AI assisted plagiarism checkers like TurnItIn and GramerlyGo

Apart from the obvious issues of originality, plagiarism, accuracy and truth, there is a major issue with copyright - "are bot-generated books and images subject to existing copyright laws"? The answer varies from country-to-country, but the broad position is currently "No", they are not protected by existing copyright and IP laws unless there is a substantial degree of human authorship (see here for a fuller discussion of this issue). In a similar vein, whilst advanced bots like ChatGPT may be acceptable in many countries, there will be some regimes that restrict their usage.

AI is also used in copyright protection services, which seek to identify infringements of copyright detected on websites (e.g. Redpoints ). In addition there are specialized services that aim to avoid accidental copyright infringement, such as those offered by the Copyright Clearance Centre (CCC) and others. Many of these AI-driven services offer obvious and immediate benefits to content authors, publishers and educationists.

A further area in which AI can assist publishers is in the secure distribution and use of electronic documents (as PDF or other format files). Some of the facilities that are already being used or are in development to enhance security include malware and virus detection (discussed at some length in our previous newsletter); identification of sensitive content (e.g. credit card or bank details); and enhanced encryption and access controls, including biometrics (although currently the latter tends to be device-specific).

ChatGPT response to the question, What is the role of AI in publishing?

"There are several ways in which AI can be applied to publishing:

  • Content Creation: AI-powered writing assistants can help authors in generating content by suggesting phrases, correcting grammar, and even generating full sentences or paragraphs. Some examples of such AI writing tools are GPT-3, Jarvis, and Quill
  • Content Curation: AI-powered algorithms can help publishers in curating content from multiple sources by analyzing user preferences and recommending relevant content. Some examples of such AI-powered curation tools are Flipboard and Feedly
  • Content Distribution: AI-powered algorithms can analyze user behavior to personalize content distribution, including email marketing campaigns, social media, and other distribution channels. This enables publishers to tailor their messaging to the interests of each individual reader
  • Editorial Decision Making: AI-powered tools can help editors in making informed decisions about which content to publish and how to optimize it for maximum engagement. AI can analyze reader behavior and trends to suggest topics, headlines, and keywords for maximum reach and impact
  • Metadata Optimization: AI-powered tools can help publishers in optimizing metadata for each piece of content to improve discoverability and SEO. This includes analyzing title tags, descriptions, and other metadata to ensure they are optimized for search engines and social media algorithms
  • Translation: AI-powered translation tools can help publishers in translating content into multiple languages, making it accessible to a global audience. Some examples of such AI-powered translation tools are Google Translate and DeepL

Overall, AI has the potential to revolutionize the publishing industry by improving efficiency, productivity, and user engagement"

Software updates

Javelin3 for Windows (March/April updates):

  • Now handles password-protected standard PDF files (prompts for password before allowing access)
  • Asks users for confirmation before opening an external link in a document (for more details, see our PDF Security News #3/2023)
  • Enhancements to existing security features for DRM-protected PDF files (details available to 'bona fide' publishers, on request)
  • Release notes are now only on the Release Notes page - a new link is provided under the Javelin3 for Windows "Help" menu
  • Minor User Interface changes to catalog thumbnail handling

Javelin3 for MacOS:

  • Enhancements to existing security features for DRM-protected PDF files (details available to 'bona fide' publishers, on request)


  • Major enhancements and improvements to command line/batch operation (see documentation, page 24-25 for details). The DrumlinPublisher command line interface (CLI) enables the creation of secure PDF files (drmz files) and associated authorization codes, programmatically. Essentially this is a form of API for the Drumlin DRM service. A more advanced bespoke API for .NET can also be provided, tailored to client requirements
  • Now uses more appropriate defaults for newly registered users (no expiry, unlimited views, no printing)
  • Free codes are only provided for users who complete the registration process by responding to the activation email or by email request to the Drumlin Security team

Taxbooks X for Android: Rebuilt for latest Google API level