November 14, 2026

What happened

Experts told Parliament’s Communications and Digital Committee that major technology companies are doing enough to show some compliance with the UK Online Safety Act but are not meeting the needs of individuals and young victims it is intended to protect. [1] [2]

Clare McGlynn said search engines inconsistently downrank harmful sexual-content terms: “rape porn” and “force porn” were heavily downranked, while similar terms including “unwilling porn” and “forced porn” appeared substantially higher in results. [3] [4] [5]

Refuge data cited by Janaya Walker indicated that the success rate for reports made by Refuge’s Trusted Flagger team fell from 53.7% before November 2025 to 18.2% over the subsequent 12 months; reports made to Meta had a reported 0% success rate since November 2025. [12] [13] [14] [15] [16] [17]

McGlynn cited British Columbia’s swift online civil-redress system as an example that can let individuals obtain orders removing material and, in some cases, compensation, contrasting it with the absence of an equivalent right described in the UK discussion. [23] [24] [25] [26]

Why it matters

Whiston-Dew said some publications openly advertising illegal content appear to be missed by search engines and other businesses in the publication chain, while McGlynn said extreme pornography can remain available on sites serving less regulated territories and be accessed through VPNs that bypass age assurance. [6] [7] [8] [9] [10] [11]

The article reports broader concern that the Online Safety Act has had limited impact more than a year after its duties came into force, including the Children’s Commissioner’s statement that it had made “absolutely no difference” to children’s access to harmful content. [21] [22]

Known limitations

The article cautions that Refuge’s figures should be scrutinized because Meta’s enforcement report lacks a category specifically covering violence-against-women-and-girls reports; Meta also did not immediately respond, and Google, Microsoft and TikTok did not respond to comment requests. [18] [19] [20]

Cited source evidence

These excerpts support the numbered claims above. They come from one source article and do not establish independent corroboration.

1. [1]

   Technology companies are doing just enough to show a degree of compliance with the UK's Online Safety Act, but are failing to meet the needs of the individuals and young victims the legislation exists to protect, experts say.

2. [2]

   A panel of online safety pros assembled before Parliament's Communications and Digital Committee on Tuesday, taking aim at the likes of Google, Microsoft, Meta, and ByteDance for failing to act on important matters.

3. [3]

   Clare McGlynn, professor of law at Durham University and expert on violence against women and girls (VAWG), was especially critical of major search engine providers for their inconsistent downranking of harmful content search terms.

4. [4]

   "Rape porn," for example, is correctly downranked and no user sees any relevant links until the tenth page of results, although adjacent terms such as "unwilling porn" can be found as early as the third page, McGlynn said.

5. [5]

   "So, they don't rank some of them, but just not systematically or wholly consistently." McGlynn went on to say that during her own research, "force porn" was rightly downranked heavily, although terms as similar as "forced porn" were not.

6. [6]

   Alexandra Whiston-Dew, partner at law firm Miscon de Raya, argued that search engines have such an important role in regulating the OSA, noting the limited resources at Ofcom's disposal.

7. [7]

   She said that more needs to be done from the tech side to ensure different parties in the chain of publication are detected early and downranked or blocked accordingly.

8. [8]

   There are publications, for example, that actively advertise themselves as websites that host illegal content, and these appear to be missing from the search engines' crosshairs.

9. [9]

   "So, they are spotlighting the fact that they are publishing illegal content, and this is going under the radar not only from the search engines but also the other platforms and the businesses in the chain of publication," said Whiston-Dew.

10. [10]

    "Also Ofcom needs to really pull all of their resources together and attack that receptacle of illegal content." Similarly, McGlynn raised the issue of porn sites continuing to host extreme porn, as they are legally allowed to do, to make it available to less regulated territories.

11. [11]

    With the harmful content still on the site, all it takes is a VPN connection to access it, and bypass the age assurance mechanisms in the process.

12. [12]

    Success of user reports trails away As well as a piecemeal approach to downranking extreme content, experts said that tech platforms initially showed a strong appetite in quickly handling content-takedown requests, although this appears to have waned substantially.

13. [13]

    Citing data from domestic violence and VAWG charity Refuge [PDF], Janaya Walker, interim director at the End Violence Against Women Coalition (EVAW), said that before November 2025, the success rate for these kinds of reports made using Refuge’s Trusted Flagger team stood at 53.7 percent.

14. [14]

    Over the past 12 months, however, this dropped to just 18.2 percent.

15. [15]

    Further, since November 2025, zero percent of reports made to Meta have been successful, according to Refuge’s figures.

16. [16]

    The charity’s Trusted Flagger team works closely with social media platforms to report content that victims may have reported to it.

17. [17]

    The idea is that if the report comes from an authority like Refuge, it is acted upon more quickly.

18. [18]

    Meta did not immediately respond to questions about its takedown rate, but according to its Community Standards Enforcement Report, it removed 27.8 million posts related to adult nudity and sexual activity between April and June 2026.

19. [19]

    The report does not include a category devoted to VAWG-related reports, so take the data with the appropriate level of scrutiny.

20. [20]

    Neither Google, Microsoft, nor TikTok responded to requests for comment.

21. [21]

    The findings from Refuge and the discussion in Parliament this week reflect the wider concern about the efficacy of the OSA, and how online safety experts are frustrated with how little impact it has had, more than a year after the new duties came into force.

22. [22]

    England's Children's Commissioner, Dame Rachel de Souza, told the same committee in September that the OSA "has made absolutely no difference" to children's access to harmful content online, as evidenced by the accounts of young people themselves.

23. [23]

    Amid discussions from the panel of experts this week regarding the need for a civil redress scheme, McGlynn pointed to British Columbia's system as an example of how other countries are going further to protect people from online harms.

24. [24]

    She said: "...an individual still does not have the right to demand an image be taken down.

25. [25]

    And that is where I think we just lack that capacity to say to an individual, 'this is what you need to do, and you can bring this legal action, for example, against X, against the perpetrator, you can get the material deleted.' "In British Columbia in Canada, they have a swift online civil redress system where you can make these claims, and you can get instant orders to get this material taken down, and you can even get compensation for some of the harm suffered.

26. [26]

    "So other countries are doing more, if you like, to give the individuals that right as well as all the other measures at the platform level." ®

Read the original article →

What happened

CISA added a high-severity UnrealIRCd security flaw to its Known Exploited Vulnerabilities catalog after reports of active exploitation. [1]

The flaw is tracked as CVE-2026-114713 and has a CVSS score of 8.6. [2]

The vulnerability is described as improper neutralization of delimiters. [2]

Cited source evidence

These excerpts support the numbered claims above. They come from one source article and do not establish independent corroboration.

1. [1]

   Cybersecurity and Infrastructure Security Agency (CISA) on Friday added a high-severity security flaw impacting UnrealIRCd, an open-source Internet Relay Chat (IRC) server, to its Known Exploited Vulnerabilities (KEV) catalog, following reports of active exploitation.

2. [2]

   The vulnerability, tracked as CVE-2026-114713 (CVSS score: 8.6), is a case of improper neutralization of delimiters

Read the original article →