Yellow abstract of floor level.

What the AI is going on… December 2023 to March 2024

Published on 17 April 2024

December 2023

AI in the palm of your hand

Google launched its newest GenAI, Gemini Nano. Designed specifically for mobile phones it presents new competition to market leaders OpenAI. Gemini Nano will allow AI apps that operate offline as opposed to only on servers and provide improved privacy for users.

Dealing in news

News organisation Axel Springer, whose brands include Politico and Business Insider, has made a deal with OpenAI to allow its GenAI chatbot, ChatGPT, to use the company’s news content. As part of the groundbreaking deal, ChatGPT will use Axel Springer’s news content to train the AI model and will provide ChatGPT users with news story summaries and a link to the original article.

Light-touch UK

Michelle Donelan, Secretary of State for Science, Innovation and Technology told the House of Commons Science, Innovation and Technology Committee that the UK will not prematurely legislate on AI. The UK’s pro-innovation approach has faced scrutiny in recent months as other countries, and in particular, the EU look to implement robust statutory measures in response to the AI boom. For now, it appears the UK remains steadfast in its desire to appear tech-friendly and will focus on the role of regulators.

January 2024

UK Gov sets new test for laws on AI

The UK Government continued its “pro-innovation” approach to regulating AI, once again resisting calls for tougher regulation. The Government is expected to publish a set of criteria which must be met before rules can be placed onto big players such as ChatGPT and Bard. The criteria will form part form part of the Government’s response to the white paper, A Pro-innovation approach to AI regulation, published for consultation in March 2023.

Apple’s AI Acquisitions

Apple made moves to enhance its AI offering – an answer to Google’s Gemini Nano? The high-profile acquisitions include a voice recognition company, a facial recognition startup and a machine learning platform, alongside the hiring of new talent in the AI field.

February 2024

Government distracted by doomsday AI risks

The House of Lords Communications and Digital Committee published its report on an inquiry into LLMs and generative AI. The inquiry focused on identifying and assessing the real potential risks of LLMs and recommendations for the UK Government to appropriately balance risk and innovation. The Committee was particularly concerned with the Government’s “narrow view of high‑stakes AI safety” and encouraged the Government to focus on framing LLM innovation in a more “positive” light.

Google lands Reddit

Reddit has agreed to license the content posted on its platform by users to Google for a reported $60 million a year. Google will use the data to train its AI tools. This deal is the latest in a string of agreements being reached between social media platforms and AI companies.

March

Stability AI defends its position

Getty Images’ claim against Stability AI for infringement of copyright, trade marks and database rights continued this month, as Stability AI filed its defence in the case. To assert its position that its output would not constitute an infringement, Stability AI claims that any outputs of Getty work were an “infinitesimally small part of the expression” of Getty and the copyright holders, and consequently “it cannot be said that any reproduction derives from the copyright work”. A key part of Stability AI’s defence is that it cannot be liable for copyright infringement in the UK because it did not train the model in the UK. The case continues...

Spring 2024