It may be a new year, but it is certainly not a new story for AI with another bumper issue round-up of the stories that have caught our eye. December delivered a blizzard of regulatory and commercial moves, from the European Commission’s antitrust probe into Meta’s alleged exclusionary use of its AI assistant on WhatsApp, to Disney’s headline‑grabbing $1 billion bet on OpenAI to bring its characters to Sora, while public markets gave many a rocky ride into the year end with AI barometers such as Nvidia, Oracle, AMD and Broadcom all seeing downward stock movement, amid broader AI sector volatility into year-end.
UK developments
DSIT reports progress on copyright works in AI development
The UK government’s consultation on potential changes to copyright law ended in February last year. December saw the Department for Science, Innovation and Technology (DSIT), the Intellectual Property Office (IPO) and the Department for Culture, Media and Sport (DCMS) publish their statement of progress as required under section 137 of the Data (Use and Access) Act 2025. The report outlines progress for an economic impact assessment and a report on copyright works in AI development, both due before Parliament by 18 March 2026. It also extends consideration to AI systems developed overseas, a point underscored by Getty Images v Stability AI, where Getty abandoned its' primary infringement claim over Stable Diffusion's foreign training but secured permission to appeal the High Court's rejection of secondary infringement claims for importing, possessing, selling, or distributing the model in the UK.
Parliamentary Conference on AI addresses oversight and rights
The Parliamentary Conference on Artificial Intelligence, organised by the Parliamentary Assembly of the Council of Europe (PACE) and the UK Parliament, convened at Westminster on 15-16 December. Sessions covered AI's alignment with human rights and democratic values, risks in democratic processes like misinformation, parliamentary oversight mechanisms, and policy approaches to regulate AI effectively. Participants included MPs from across Europe, policymakers, AI experts, academics, and NGO representatives, with discussions emphasising parliaments' proactive role in shaping AI governance amid rapid technological evolution.
DSIT partners with DeepMind to boost AI innovation
DSIT announced a new partnership with Google DeepMind to accelerate the use of AI across public services, education, and scientific research. Under a voluntary memorandum of understanding, DeepMind will open its first automated research lab in the UK in 2026, focusing on AI and robotics-driven superconductor materials research using Gemini technology. The collaboration also explores developing a “Gemini for Government” platform to automate civil service tasks and a curriculum-aligned Gemini model for education. In addition, UK scientists will gain priority access to DeepMind’s leading AI tools, including AlphaGenome and AlphaFold.
Home Office consults on AI facial recognition framework
The Home Office launched a consultation to establish a new legal framework for law enforcement's use of AI-driven facial recognition and related technologies, seeking views on covered tools (e.g., biometric, inferential, and object recognition systems), applicable organisations, and safeguards to balance privacy with proportionate deployment. Running until 12 February, the consultation addresses ongoing debates over real-time and retrospective systems amid judicial scrutiny. This builds on prior court challenges (e.g. R (Bridges) v Chief Constable of South Wales Police), reinforcing calls for statutory clarity in public authority AI use.
In the European Union…
AI Board meeting
The AI Board met to discuss the Digital Simplification Package and review current priorities for implementing the AI Act.
EU launches secure whistleblower tool
The Commission announced a confidential whistleblower platform allowing individuals to report suspected breaches of the AI Act directly to the EU AI Office. The tool enables users to submit anonymous reports in any EU language, supported by secure encryption, and to track the progress of their reports. This initiative aims to strengthen enforcement, increase early risk detection, and uphold transparency and trust in AI development across Europe, with statutory protections against retaliation coming into effect from August 2026, pursuant to the EU Whistleblowing Directive.
EU approves amendment to create AI gigafactories
The European Council agreed on an amendment to the regulation that sets the framework for the activities of the European High-Performance Computing Joint Undertaking. The amendment will establish a new funding and procurement framework, adds protections for start-ups and scale-ups, and allows unused EU funds to be redirected to support AI development projects.
Commission launches consultation on TDM opt-outs
The Commission opened a consultation on protocols for text and data mining under the AI Act and the GPAI Code of Practice to support implementation of the requirement for general-purpose AI model providers to identify and comply with reservations of rights expressed by rightsholders. Running until 23 January, it seeks views on technical feasibility and uptake of text and data mining (TDM) opt-out solutions and invites expressions of interest for follow-up workshops to discuss opt-out protocols that GPAI Code of Practice signatories must respect under the AI Act. At its conclusion, the Commission will publish a list of generally agreed machine-readable opt-out solutions, to be reviewed at least every two years alongside Code of Practice updates.
Meta comes under investigation…again
The Commission opened an investigation into Meta’s embedding its own AI assistant into WhatsApp while implementing policy changes which effectively ban rival AI chatbots from the messaging app. The formal antitrust investigation will look into Meta’s new policy (due to take effect on 15 January) for existing AI providers, although new providers have been blocked since October last year. The investigation stems from complaints by smaller companies, including The Interaction Company of California, which developed the AI assistant Poke.com, and Spanish startup Luzia.
As does Google…
The Commission also launched a formal antitrust investigation into whether Google violated EU competition rules by using web and YouTube content for AI purposes without fair compensation or opt-out options. The inquiry will assess whether Google imposed unfair terms on publishers and creators, such as using their material in AI Overviews and AI Mode on search results, or requiring YouTube uploaders to allow their videos to train generative AI models without remuneration or refusal rights. The Commission will also examine whether Google gave itself preferential access to online content, disadvantaging rival AI developers, which could amount to an abuse of dominance under Article 102 TFEU and Article 54 of the EEA Agreement.
AI Liability Directive canned
As previously reported, the draft AI Liability Directive (AILD) is no more. On 3 December the European Parliament’s Legal Affairs Committee (JURI) rejected a proposal to challenge the Commission's withdrawal of the AI Liability Directive. The vote in the committee was overwhelmingly against the challenge, with 22 votes opposed and only one in favour. The JURI’s decision effectively puts an end to the legislative process for the AILD.
Innatera and 42 Technology partner-up on neuromorphic processors
Netherlands-based Innatera announced a partnership with UK consultancy 42 Technology to accelerate the deployment of its neuromorphic processors into commercial products. The collaboration aims to integrate Innatera's brain-inspired Pulsar chip - which processes sensor data for less than one milliwatt - into intelligent devices ranging from consumer electronics to industrial systems. Neuromorphic processors mimic the human brain's neural architecture with potential to slash energy consumption of AI systems by up to 1,000 times compared to conventional chips.
Deutsche Telekom and Schwarz Group plan German AI ‘Gigafactory’
Deutsche Telekom and the Schwarz Group (owner of Lidl and Kaufland) are reported to be in advanced talks to jointly build a large-scale AI data centre in Germany. The gigafactory is designed to support AI’s massive computational needs, with both companies aiming to secure funding from the EU’s InvestAI initiative to close the continent’s compute gap with the US and China. The partnership aims to leverage Deutsche Telekom’s telecom and cloud expertise alongside Schwarz Group’s retail scale and land resources.
In other news…
Turkey’s KVKK publishes generative AI and personal data protection guide
The Data Protection Authority (KVKK) released a Generative Artificial Intelligence and Personal Data Protection Guide (in 15 Questions), which clarifies how generative AI systems are regulated under Turkey’s Personal Data Protection Law. The guide explains key concepts such as deepfakes and neural networks, and provides practical advice for data controllers on compliance, including the risks of sharing personal data with AI systems, the importance of anonymisation, and the need for transparency and human oversight in automated decision-making. The document also emphasises ethical use, data minimisation, and the necessity of reviewing AI tools’ data policies to ensure alignment with Turkish and international privacy standards.
Chinese AI models surpass US in global downloads
Chinese AI models overtook American developers in global downloads for the first time, capturing 17% of the market compared to 15.8% for US creators, according to a joint study by the Massachusetts Institute of Technology and Hugging Face. The shift was driven largely by DeepSeek and Alibaba's Qwen models which have gained traction by offering cost-effective alternatives with frequent updates. Separately it was reported that DeepSeek is developing its next major AI model using thousands of Nvidia's export controlled advanced Blackwell chips alleged to have been smuggled into China through a complex operation involving initial shipment to data centres in US export-approved countries, followed by dismantling servers and importing the components in pieces.
Meta’s Avocado shifts to closed AI
With Meta's Llama 4 open-source model having fallen short of expectations following its April 2025 release, the social media giant announced a closed AI model, codenamed Avocado, which is expected to launch in spring 2026. The move to a tightly controlled, proprietary model, which can be sold to enterprise buyers, aligns Meta with rivals like Google and OpenAI.
Digital Connexion plans data centre cluster in India
Digital Connexion, the Reliance–Brookfield–Digital Realty joint venture, is reported to be investing $11 billion over five years to build a 1 gigawatt AI-native data centre cluster in Visakhapatnam, dramatically expanding India’s hyperscale and AI compute capabilities. The facility, spanning 400 acres, will be purpose-built for high-density AI workloads and next-generation cooling.
McKinsey takes a dose of its own medicine
Consulting giant McKinsey was reported to have held initial talks with the heads of non-client-facing departments about shrinking their teams by as much as 10 per cent. Several thousand jobs are said to be at risk over the next couple of years, as the company seeks to slim down non-client facing support roles, whilst leveraging “new technology” such as AI and automation.
Disney’s about turn
The Walt Disney Company announced a three-year licensing agreement with OpenAI. The deal sees Disney become the first major content partner on Sora, OpenAI's AI video generation platform, enabling users to create short videos featuring more than 200 Disney, Marvel, Pixar and Star Wars characters, including Mickey Mouse, Darth Vader, Iron Man, and Elsa, although strict end user guardrails will be put in place. Under the deal, Disney will invest $1 billion in equity and receive warrants to purchase additional shares in the AI company. The partnership marks an about turn for the entertainment giant - a matter of months ago Disney demanded OpenAI remove all of its intellectual property from Sora.
If you’d like to chat about what these developments could mean for your business, feel free to get in touch with Tim Wright and Nathan Evans.
