Owen Bennett

Interesting scoop by @danielthomasldn.bsky.social on steps taken by YouTube to frustrate third-party audience measurement on smart TVs.

Under EU rules, media plurality is a systemic risk for which YT must facilitate independent research. Might offer a pathway through.

www.ft.com/content/9493...

Many UK Users Soon Won't Be Able to Access Pornhub Starting February 2, many people connecting from the UK will not be able to access the porn site and many others.

The story in @404media.co here www.404media.co/uk-pornhub-b...

It's worth noting that Ofcom's implementing guidance affords adult companies significant flexibility in how to deploy age assurance safety and efficiently on their service. The fact that that wasn't enough to motivate the company to comply through time is a worrying omen.

The company has for years campaigned for global age assurance requirements to be levelled against operating system and app store operators (i.e. Google and Apple), and not websites like theirs. This development is a significant escalation in that global campaign.

Aylo, the company behind Pornhub and other leading adult websites, will soon restrict access to the service for most of its UK users, so as not to have to comply with the Online Safety Act's age assurance duties.

Everything @jamesrball.com says below, as well as the backsliding precedent it sets among open societies for VPN restrictions. An often underappreciated bulwark against banning VPNs in open societies is the association risk (look who you're associating with globally). This would break that bulwark.

@saraheskens.bsky.social has a good paper on this in the latest issue of JIPTEC (open access). www.jipitec.eu/jipitec/arti...

What's most interesting about this development is what it reveals about the frontiers, or rather, gaps, in the EU's platform regulation rulebook. It feels odd to bifurcate WhatsApp-like services along a 'private'/'public' dichotomy (e.g. when group chats on WhatsApp can hold over 1000 participants).

The European Commission has today - as expected - designated WhatsApp as a Very Large Online Platform under the DSA, with respect to its 'Channels' functionality. Crucially, instant messaging and group chat functionalities are excluded. digital-strategy.ec.europa.eu/en/news/comm...

Commission investigates Grok and X\'s recommender systems under the Digital Services Act The European Commission has launched a new formal investigation against X under the Digital Services Act (DSA).

Confirmed - EU Commission opens investigation into X as regards Grok nudifications - and also widens its earlier investigation into X to include recommender systems - ec.europa.eu/commission/p...

Background in context of the previous fine against X here: eulawanalysis.blogspot.com/2025/12/the-...

Behavioural regulatory frameworks that incorporate ongoing supervision (i.e. the DSA and OSA) foresee this informal coordination between regulator and company as the primary mechanism for achieving regulatory outcomes. Let's hope for continued (and ever-deeper) transparency in how it plays out.

The European Commission's opening of an investigation into xAI is naturally going to get a lot of attention today. In another eye-catching enforcement-related update, the UK's Ofcom has seemingly extracted voluntary concessions from Snap regarding its management of online safety risks.

Exhibit n (via @wired.com) on why policy proposals starting from the premise of 'digital sovereignty' and 'strategic autonomy' should never get a free pass in policymaking circles, especially when they involve the shifting of ownership/control between private operators. www.wired.com/story/tiktok...

As age assurance becomes a more entrenched feature of the global internet experience in 2026, the risk of trust-busting data protection breaches will invariably rise. Good to see this affirmation; let's hope it's backed up with robust supervision of age assurance deployments.

The regulators' principles for age assurance deployments map on closely to existing guidance concerning the UK OSA (Part V) and the EU DSA (art. 28). These approaches appear to be emerging as a de facto global norm.

Regulators are eager to demonstrate to the large industry players (particularly in the adult sector) that their compliance pays. For many of the largest adult services, opposition to age assurance stems largely from fears of distortive effects on the market arising from patchy enforcement.

The paper reflects an emerging consensus amongst lawmakers and independent regulators in different jurisdictions regarding the inevitability of age assurance, in situations where the policy objective is age-appropriate experience or access bans for children.

Interesting development for age assurance watchers: The Global Online Safety Regulators Network has issued a new position statement, setting out regulators' attitude to age assurance as a regulatory tool. Some initial reflections in thread - www.cnam.ie/app/uploads/...

p.s. These macro-trends overlay a bunch of additional regulatory questions that the field will grapple with in 2026 (not to mention spillover themes like digital sovereignty). If you’re working on these topics and you’d like to discuss them, I’m always happy to chat 🤝

💰 2025 saw legal challenges to supervisory fees in the EU and Ireland. With Ofcom beginning to charge fees in the UK this year, and the AU government signalling a possibility of industry fees in the upcoming revision of its rulebook, legal fights about the regulatory bill will surely heat up. (9/9)

💰 The (monetary) cost of regulation: Compliance costs money. It always has and always will. Where we’re likely to see some novelty on this topic is that it will focalise around the supervisory fees that companies have to increasingly pay for the privilege of being regulated. (8/9)

⚠️ The X/Grok scandal is unique in some respects (E**n, ) but in others (e.g. rapid integration of new GenAI features with inadequate risk assessment and mitigation), it could well be a harbinger for a serious systemic challenge facing regulatory supervisors in 2026. (7/9)

⚠️ We can expect an accelerating race this year by established platforms to integrate GenAI applications and features into their products, and with it, the risk of more serious safety failures for users. (6/9)

⚠️ AI bubble on a collision course with safety-by-design: The theory of change underpinning today’s online safety laws is one of ‘procedural accountability’ (a form of regulated self-regulation). This regulatory philosophy for online safety wasn’t developed with the AI bubble in mind. (5/9)

🚸 Should the EU/UK go down this route regulators and companies in those jurisdictions will be in a tricky position - the DSA and the OSA aim to make social media age-appropriate for children, and regulators and social media companies have invested heavily to that end. Will it all be for naught?(4/9)

🚸 The number of jurisdictions opting for social media bans will increase significantly later this year. In the eyes of lawmakers, bans are easy to communicate to the public and their impact on online experience is immediate. They are also immensely popular with voters. (3/9)

🚸 Children and social media at a crossroads: 2026 will see a divergence between those jurisdictions that seek to make social media age-appropriate for children, and those that consider social media inherently unsafe for children (and worthy of banning). (2/9)

If the first few weeks are anything to go by, this year is likely to be another one of whirlwind developments for those of us working on online safety regulation globally.

Amongst all the messy data points, there’s some overarching policy trends on which I’ll be keeping a keen eye (1/9):

Urban mobility is amongst the lowest hanging fruits in the race to net zero. That we can't maintain the public policy environment that facilitates net zero in urban mobility (in this case, practices that reduce the incentive for individual car ownership) is a travesty.

Europe prides itself on an ecosystem of independent comms regulators. However those regulators choose to address the compliance failures of X (and assuming their approach adheres to the rule of law and procedural rules), they'll need the backing of their governments to weather the geopolitical storm