Government & Policy

EU plan to force messaging apps to scan for CSAM risks millions of false positives, experts warn

Comment

Image Credits: Bryce Durbin / TechCrunch

A controversial push by European Union lawmakers to legally require messaging platforms to scan citizens’ private communications for child sexual abuse material (CSAM) could lead to millions of false positives per day, hundreds of security and privacy experts warned in an open letter Thursday.

Concern over the EU proposal has been building since the Commission proposed the CSAM-scanning plan two years ago — with independent experts, lawmakers across the European Parliament and even the bloc’s own Data Protection Supervisor among those sounding the alarm.

The EU proposal would not only require messaging platforms that receive a CSAM detection order to scan for known CSAM, but they would also have to use unspecified detection scanning technologies to try to pick up unknown CSAM and identify grooming activity as it’s taking place — leading to accusations of lawmakers indulging in magical thinking-levels of technosolutionism.

Critics argue the proposal asks the technologically impossible and will not achieve the stated aim of protecting children from abuse. Instead, they say, it will wreak havoc on internet security and web users’ privacy by forcing platforms to deploy blanket surveillance of all their users in deploying risky, unproven technologies, such as client-side scanning.

Experts say there is no technology capable of achieving what the law demands without causing far more harm than good. Yet the EU is plowing on regardless.

EU watchdog questions secrecy around lawmakers’ encryption-breaking CSAM scanning proposal

The latest open letter addresses amendments to the draft CSAM-scanning regulation recently proposed by the European Council, which the signatories argue fail to address fundamental flaws with the plan.

Signatories to the letter — numbering 270 at the time of writing — include hundreds of academics, including well-known security experts such as professor Bruce Schneier of Harvard Kennedy School and Dr. Matthew D. Green of Johns Hopkins University, along with a handful of researchers working for tech companies such as IBM, Intel and Microsoft.

An earlier open letter (last July), signed by 465 academics, warned the detection technologies the legislation proposal hinges on forcing platforms to adopt are “deeply flawed and vulnerable to attacks” and would lead to a significant weakening of the vital protections provided by end-to-end encrypted (E2EE) communications.

Little traction for counterproposals

Last fall, MEPs in the European Parliament united to push back with a substantially revised approach — which would limit scanning to individuals and groups who are already suspected of child sexual abuse; limit it to known and unknown CSAM, removing the requirement to scan for grooming; and remove any risks to E2EE by limiting it to platforms that are not end-to-end-encrypted. But the European Council, the other co-legislative body involved in EU lawmaking, has yet to take a position on the matter, and where it lands will influence the final shape of the law.

The latest amendment on the table was put out by the Belgian Council presidency in March, which is leading discussions on behalf of representatives of EU Member States’ governments. But in the open letter the experts warn this proposal still fails to tackle fundamental flaws baked into the Commission approach, arguing that the revisions still create “unprecedented capabilities for surveillance and control of Internet users” and would “undermine … a secure digital future for our society and can have enormous consequences for democratic processes in Europe and beyond.”

Tweaks up for discussion in the amended Council proposal include a suggestion that detection orders can be more targeted by applying risk categorization and risk mitigation measures, and cybersecurity and encryption can be protected by ensuring platforms are not obliged to create access to decrypted data and by having detection technologies vetted. But the 270 experts suggest this amounts to fiddling around the edges of a security and privacy disaster.

From a “technical standpoint, to be effective, this new proposal will also completely undermine communications and systems security,” they warn. While relying on “flawed detection technology” to determine cases of interest in order for more targeted detection orders to be sent won’t reduce the risk of the law ushering in a dystopian era of “massive surveillance” of web users’ messages, in their analysis.

The letter also tackles a proposal by the Council to limit the risk of false positives by defining a “person of interest” as a user who has already shared CSAM or attempted to groom a child — which it’s envisaged would be done via an automated assessment, such as waiting for one hit for known CSAM or two for unknown CSAM/grooming before the user is officially detected as a suspect and reported to the EU Centre, which would handle CSAM reports.

Billions of users, millions of false positives

The experts warn this approach is still likely to lead to vast numbers of false alarms.

“The number of false positives due to detection errors is highly unlikely to be significantly reduced unless the number of repetitions is so large that the detection stops being effective. Given the large amount of messages sent in these platforms (in the order of billions), one can expect a very large amount of false alarms (in the order of millions),” they write, pointing out that the platforms likely to end up slapped with a detection order can have millions or even billions of users, such as Meta-owned WhatsApp.

“Given that there has not been any public information on the performance of the detectors that could be used in practice, let us imagine we would have a detector for CSAM and grooming, as stated in the proposal, with just a 0.1% False Positive rate (i.e., one in a thousand times, it incorrectly classifies non-CSAM as CSAM), which is much lower than any currently known detector.

“Given that WhatsApp users send 140 billion messages per day, even if only 1 in hundred would be a message tested by such detectors, there would be 1.4 million false positives every single day. To get the false positives down to the hundreds, statistically one would have to identify at least 5 repetitions using different, statistically independent images or detectors. And this is only for WhatsApp — if we consider other messaging platforms, including email, the number of necessary repetitions would grow significantly to the point of not effectively reducing the CSAM sharing capabilities.”

Another Council proposal to limit detection orders to messaging apps deemed “high-risk” is a useless revision, in the signatories’ view, as they argue it’ll likely still “indiscriminately affect a massive number of people.” Here they point out that only standard features, such as image sharing and text chat, are required for the exchange of CSAM — features that are widely supported by many service providers, meaning a high risk categorization will “undoubtedly impact many services.”

They also point out that adoption of E2EE is increasing, which they suggest will increase the likelihood of services that roll it out being categorized as high risk. “This number may further increase with the interoperability requirements introduced by the Digital Markets Act that will result in messages flowing between low-risk and high-risk services. As a result, almost all services could be classified as high risk,” they argue. (NB: Message interoperability is a core plank of the EU’s DMA.)

A backdoor for the backdoor

As for safeguarding encryption, the letter reiterates the message that security and privacy experts have been repeatedly yelling at lawmakers for years now: “Detection in end-to-end encrypted services by definition undermines encryption protection.”

“The new proposal has as one of its goals to ‘protect cyber security and encrypted data, while keeping services using end-to-end encryption within the scope of detection orders.’ As we have explained before, this is an oxymoron,” they emphasize. “The protection given by end-to-end encryption implies that no one other than the intended recipient of a communication should be able to learn any information about the content of such communication. Enabling detection capabilities, whether for encrypted data or for data before it is encrypted, violates the very definition of confidentiality provided by end-to-end encryption.”

In recent weeks police chiefs across Europe have penned their own joint statement — raising concerns about the expansion of E2EE and calling for platforms to design their security systems in such as way that they can still identify illegal activity and send reports on message content to law enforcement.

The intervention is widely seen as an attempt to put pressure on lawmakers to pass laws like the CSAM-scanning regulation.

Police chiefs deny they’re calling for encryption to be backdoored but they haven’t explained exactly which technical solutions they do want platforms to adopt to enable the sought for “lawful access.” Squaring that circle puts a very wonky-shaped ball back in lawmakers’ court.

If the EU continues down the current road — so assuming the Council fails to change course, as MEPs have urged it to — the consequences will be “catastrophic,” the letter’s signatories go on to warn. “It sets a precedent for filtering the Internet, and prevents people from using some of the few tools available to protect their right to a private life in the digital space; it will have a chilling effect, in particular to teenagers who heavily rely on online services for their interactions. It will change how digital services are used around the world and is likely to negatively affect democracies across the globe.”

An EU source close to the Council was unable to provide insight on current discussions between Member States but noted there’s a working party meeting on May 8 where they confirmed the proposal for a regulation to combat child sexual abuse will be discussed.

Europe’s CSAM-scanning plan is a tipping point for democratic rights, experts warn

European police chiefs target E2EE in latest demand for ‘lawful access’

More TechCrunch

A new crop of early-stage startups — along with some recent VC investments — illustrates a niche emerging in the autonomous vehicle technology sector. Unlike the companies bringing robotaxis to…

VCs and the military are fueling self-driving startups that don’t need roads

When the founders of Sagetap, Sahil Khanna and Kevin Hughes, started working at early-stage enterprise software startups, they were surprised to find that the companies they worked at were trying…

Deal Dive: Sagetap looks to bring enterprise software sales into the 21st century

Keeping up with an industry as fast-moving as AI is a tall order. So until an AI can do it for you, here’s a handy roundup of recent stories in the world…

This Week in AI: OpenAI moves away from safety

After Apple loosened its App Store guidelines to permit game emulators, the retro game emulator Delta — an app 10 years in the making — hit the top of the…

Adobe comes after indie game emulator Delta for copying its logo

Meta is once again taking on its competitors by developing a feature that borrows concepts from others — in this case, BeReal and Snapchat. The company is developing a feature…

Meta’s latest experiment borrows from BeReal’s and Snapchat’s core ideas

Welcome to Startups Weekly! We’ve been drowning in AI news this week, with Google’s I/O setting the pace. And Elon Musk rages against the machine.

Startups Weekly: It’s the dawning of the age of AI — plus,  Musk is raging against the machine

IndieBio’s Bay Area incubator is about to debut its 15th cohort of biotech startups. We took special note of a few, which were making some major, bordering on ludicrous, claims…

IndieBio’s SF incubator lineup is making some wild biotech promises

YouTube TV has announced that its multiview feature for watching four streams at once is now available on Android phones and tablets. The Android launch comes two months after YouTube…

YouTube TV’s ‘multiview’ feature is now available on Android phones and tablets

Featured Article

Two Santa Cruz students uncover security bug that could let millions do their laundry for free

CSC ServiceWorks provides laundry machines to thousands of residential homes and universities, but the company ignored requests to fix a security bug.

23 hours ago
Two Santa Cruz students uncover security bug that could let millions do their laundry for free

OpenAI’s Superalignment team, responsible for developing ways to govern and steer “superintelligent” AI systems, was promised 20% of the company’s compute resources, according to a person from that team. But…

OpenAI created a team to control ‘superintelligent’ AI — then let it wither, source says

TechCrunch Disrupt 2024 is just around the corner, and the buzz is palpable. But what if we told you there’s a chance for you to not just attend, but also…

Harness the TechCrunch Effect: Host a Side Event at Disrupt 2024

Decks are all about telling a compelling story and Goodcarbon does a good job on that front. But there’s important information missing too.

Pitch Deck Teardown: Goodcarbon’s $5.5M seed deck

Slack is making it difficult for its customers if they want the company to stop using its data for model training.

Slack under attack over sneaky AI training policy

A Texas-based company that provides health insurance and benefit plans disclosed a data breach affecting almost 2.5 million people, some of whom had their Social Security number stolen. WebTPA said…

Healthcare company WebTPA discloses breach affecting 2.5 million people

Featured Article

Microsoft dodges UK antitrust scrutiny over its Mistral AI stake

Microsoft won’t be facing antitrust scrutiny in the U.K. over its recent investment into French AI startup Mistral AI.

1 day ago
Microsoft dodges UK antitrust scrutiny over its Mistral AI stake

Ember has partnered with HSBC in the U.K. so that the bank’s business customers can access Ember’s services from their online accounts.

Embedded finance is still trendy as accounting automation startup Ember partners with HSBC UK

Kudos uses AI to figure out consumer spending habits so it can then provide more personalized financial advice, like maximizing rewards and utilizing credit effectively.

Kudos lands $10M for an AI smart wallet that picks the best credit card for purchases

The EU’s warning comes after Microsoft failed to respond to a legally binding request for information that focused on its generative AI tools.

EU warns Microsoft it could be fined billions over missing GenAI risk info

The prospects for troubled banking-as-a-service startup Synapse have gone from bad to worse this week after a United States Trustee filed an emergency motion on Wednesday.  The trustee is asking…

A US Trustee wants troubled fintech Synapse to be liquidated via Chapter 7 bankruptcy, cites ‘gross mismanagement’

U.K.-based Seraphim Space is spinning up its 13th accelerator program, with nine participating companies working on a range of tech from propulsion to in-space manufacturing and space situational awareness. The…

Seraphim’s latest space accelerator welcomes nine companies

OpenAI has reached a deal with Reddit to use the social news site’s data for training AI models. In a blog post on OpenAI’s press relations site, the company said…

OpenAI inks deal to train AI on Reddit data

X users will now be able to discover posts from new Communities that are trending directly from an Explore tab within the section.

X pushes more users to Communities

For Mark Zuckerberg’s 40th birthday, his wife got him a photoshoot. Zuckerberg gives the camera a sly smile as he sits amid a carefully crafted re-creation of his childhood bedroom.…

Mark Zuckerberg’s makeover: Midlife crisis or carefully crafted rebrand?

Strava announced a slew of features, including AI to weed out leaderboard cheats, a new ‘family’ subscription plan, dark mode and more.

Strava taps AI to weed out leaderboard cheats, unveils ‘family’ plan, dark mode and more

We all fall down sometimes. Astronauts are no exception. You need to be in peak physical condition for space travel, but bulky space suits and lower gravity levels can be…

Astronauts fall over. Robotic limbs can help them back up.

Microsoft will launch its custom Cobalt 100 chips to customers as a public preview at its Build conference next week, TechCrunch has learned. In an analyst briefing ahead of Build,…

Microsoft’s custom Cobalt chips will come to Azure next week

What a wild week for transportation news! It was a smorgasbord of news that seemed to touch every sector and theme in transportation.

Tesla keeps cutting jobs and the feds probe Waymo

Sony Music Group has sent letters to more than 700 tech companies and music streaming services to warn them not to use its music to train AI without explicit permission.…

Sony Music warns tech companies over ‘unauthorized’ use of its content to train AI

Winston Chi, Butter’s founder and CEO, told TechCrunch that “most parties, including our investors and us, are making money” from the exit.

GrubMarket buys Butter to give its food distribution tech an AI boost

The investor lawsuit is related to Bolt securing a $30 million personal loan to Ryan Breslow, which was later defaulted on.

Bolt founder Ryan Breslow wants to settle an investor lawsuit by returning $37 million worth of shares