Monthly Archives: April 2024

Facebook (now Meta) raised international eyebrows in 2016 when it sought to dismiss a lawsuit by a 14-year-old girl from Northern Ireland whose nude image was uploaded without her consent to a so-called “shame page.” It was a curious move by an organization that claims to take these issues seriously.  

Since then, advancements in technology have opened dangerous new doors for predators determined to weaponize social media by committing online acts of sexual violence. Yet Meta and other online platforms haven’t changed their posture; in fact, they’ve lobbied Congress to escape accountability for the behavior of bad actors on their networks.  

That 2016 case is ever-relevant today as social media companies enjoy blanket protection over the growing problem of AI “deepfakes” — images of real people transformed into lifelike pictures and videos depicting them saying and doing things that never happened.  

There was a time, not long ago, when an altered picture — “photoshopped” was the common term — could be spotted a mile away. AI’s evolution has made detecting a real image from a doctored one nearly impossible.  

Concern over AI deepfakes has largely focused on their use in perpetrating election interference in this year’s U.S. presidential race. But they raise a more depraved problem that should scare all of us: deepfake pornography, where software programs accessible by a simple online search can turn an innocent image of an unwitting individual into a sexualized scene or video that can be posted online without consent.  

Taylor Swift, along with thousands of other celebrities, was the target of one. Innocent children are being victimized by it, too. 

Not long ago a 14-year-old boy notified the National Center for Missing & Exploited Children that someone had threatened to post a deepfake pornographic image of him if he didn’t pay the perpetrator ransom.  

Another 14-year-old girl from New Jersey told the New York Times she was “summoned” to her school’s assistant principal’s office, where she was told some of her male classmates had used an AI program to turn a clothed picture of her into a naked image that was posted online.  

She recalled the shame she felt when boys started laughing at her in the hallway. Images of other girls at the school were altered and shared as well.  

One of the victims has filed a lawsuit against a male classmate allegedly involved. It’s unclear if anyone was disciplined, but it’s safe to say the boys at the school aren’t laughing anymore. 

Local, state and U.S. officials are behind the curve in addressing this problem. 

“All school districts are grappling with the challenges and impact of artificial intelligence,” read the uninspired statement of Superintendent Raymond González in response to the New Jersey incident. A recently introduced state bill spurred “dozens” of other deepfake victims to come forward.  

Social media companies can do something about it, but some simply choose not to. The New York Times’s Nicholas Kristof found that Google and Bing searches for deepfake pornographic content returned a high number of faked celebrity sex videos while Yahoo’s search engine returned none. Comparatively, a Google search on suicide returned no results on how to commit the act but rather offered listings where people can go to get help.  

“In other words,” Kristof posits, “Google is socially responsible when it wants to be, but it seems indifferent to women and girls being violated by pornographers.”   

Social media companies have proven they can’t police themselves, but Congress can. 

They can amend the Communications Decency Act to hold social media companies liable when deepfake pornographic images are published on their platforms. They can pass bills, such as the Preventing Deepfakes of Intimate Images Act and the Shield Act, which would make the circulation of deepfake pornography a crime. They can pass the Defiance Act, which would enhance deepfake pornography victims’ rights.  

But Congress has gutlessly failed to act on any of these measures.  

Social media platforms could be part of the solution. They could ally with victims and show leadership and compassion by enacting tougher measures to censor deepfake content. They could set an example by pushing for stricter laws and regulations to prevent faked sexual images from being searchable or uploaded on the networks they host.  

Instead, they seek absolution of the problem. They fight to protect themselves from liability while victims are shamed on their platforms for something they never intended, and never authorized, that will haunt them for the rest of their lives.  

Facebook lost its attempt in 2016 to have the case dismissed and wound up settling with the 14-year-old girl from Belfast. It should be a message to every other victim of revenge and deepfake pornography on these platforms: Sue them and hit them where it hurts.  

As long as Congress remains impudent in standing up to them and as long as social media companies fight tooth and nail to evade responsibility, victims should drown them in litigation.  

Maybe then, and only then, they’ll get the message that the burden is on them to solve this crisis. 

Lyndon?Haviland, DrPH, MPH, is a distinguished scholar at the CUNY School of Public Health and Health Policy.? 

Visit link:

Deepfake victims must punish Big Tech because Congress wont - The Hill

The latest craze sweeping the United States legislation to protect kids data and overall online safety has its own snappy epithet. The Guardian reports on the so-called Kids Code bills popping up in multiple state legislatures, the latest of which recently passed in Maryland by unanimous vote. The full list of nine states reads like a fellowship of age-appropriate design: Maryland, plus Vermont, Minnesota, Hawaii, Illinois, New Mexico, South Carolina, New Mexico and Nevada.

But every fellowship has its Nazgl, and in this case the two sides warring for moral control of the internet involve some atypical partnerships. Social media companies are pushing back against the legal wave alongside porn distributors and civil rights advocates, who say age verification rules risk violating the constitutional rights of law-abiding adults. For the social media firms, however, it may be less a matter of ethics and more about not wanting to enforce age policies that would limit their massive user bases all of which have been established under relatively lax verification standards.

Making matters even more complicated is the assertion by critics that social media platforms should not have to verify users age with ID or biometric verification because they already know a users age, as proven by targeted advertising. The argument is summed up tidily by a representative from the Tech Oversight Project, quoted in the Guardian: Social media companies business models are based on knowing who their users are.

In a legal petition to the Supreme Court concerning what it alleges is an unconstitutional age-verification provision in Texass HB 1181, Vera Eidelman, staff attorney with the ACLU Speech, Privacy and Technology Project, argues that the legislative panic over kids accessing content adult content is an overreaction with historical precedent.

This isnt the first time that concerns about minors access have led legislators to pass unconstitutional laws, reads the statement from Eidelman. Weve gone through this time and again, with everything from drive-in movies to video games to websites, and courts have repeatedly struck down laws imposing requirements that burden adults access to non-obscene sexual content in the name of protecting children.

Tactics employed by social media firms have not done much to dial down the tone. Lobbyists have posed as concerned parents in court without disclosing their affiliations. State disclosure forms reveal that big tech companies spent more than $243,000 in lobbying fees in Maryland in 2023, with Google spending $93,076, Amazon $88,886 and Apple $133,449. NetChoice, the industry lobby group representing the firms, has its own set of proposed solutions that would eliminate the need for identity verification, most of which take the onus to protect kids off of them. The Tech Oversight Project has observed a clear and accelerating pattern of deception in anti-Kids Code lobbying.

It makes for what John Carr, Secretary of the UKs Childrens Charities Coalition on Internet Safety (CHIS) and a noted authority on young peoples use of the internet, calls an exceptionally uneven playing field.

Speaking on the fifth and final day of the 2024 Global Age Assurance Standards Summit, Carr says NetChoice frames its mission as to defend free enterprise and free speech on the internet. Carr disagrees. The only thing that NetChoice actually does is take to court every single federal, state or city piece of legislation that tries to introduce any kind of regulation on anything connected with the Internet. And the reason for that is very straightforward they dont mind if they lose. But if they delay the process by five years, four years, six years, and the status quo is maintained, thats money in the bank.

While third-party vendors are the simplest and most accessible solution for age verification, they come with privacy risks a problem that Frances National Commission on Informatics and Liberty (CNIL) aims to solve with its more private system.

In an interview with Scientific American, Olivier Blazy, a computer scientist and professor at the cole Polytechnique in France who worked on CNILs age verification scheme, says the regulators system creates something like a firewall between the content provider and the verification service. The only information the content provider gets is a yes or no about whether a user is aged 18 or older, Blazy says. The only information the age verifier gets is that someone has sent an age-verification request.

ACLU | age verification | biometrics | children | CNIL | legislation | regulation | social media

Apr 18, 2024, 3:16 pm EDT

Compliance with biometric presentation attack detection standards has become table stakes for numerous applications of face biometrics in particular, and

Apr 18, 2024, 2:56 pm EDT

Representatives of the W3C Advisory Committee are looking for participants in a Federated Identity Working Group after a draft charter

Apr 18, 2024, 2:52 pm EDT

Digital birth certificates are going live in New South Wales, and kids will be the first to use them. The

Apr 18, 2024, 2:26 pm EDT

A method of using public key infrastructure (PKI) to enable interoperability among age assurance providers and systems was shared at

Apr 18, 2024, 12:59 pm EDT

Catholic bishops under the banner of the National Episcopal Conference of Cameroon (NECC) have launched a fervent appeal to all

Apr 18, 2024, 12:38 pm EDT

The distress of Nigerians over repeated episodes of biometric capture for different identification purposes has been highlighted by local outlet

Read the original:

Kids Code bills prompt epic showdown between regulators, activists and big tech firms - Biometric Update

Big tech firms have been solid business for office landlords across the country, but the end of a growth spurt in the sector means thats no longer the case.

Many of the top technology companies are downsizing or reevaluating their office needs across the country, the Wall Street Journal reported. Those firms reluctance to keep vast amounts of space is another hit to commercial owners.

Amazon, one of the largest office tenants in the entire country, is looking to reduce its office vacancies by passing on lease renewals, terminating some deals early and dropping underutilized floors.

Alphabets Google and Meta are also reconsidering their office needs. The former has listed office space in tech mecca Silicon Valley for sublease, while Mark Zuckerbergs company is leasing less space than it did at the onset of the pandemic, an acknowledgement that remote work is here to stay, as well as a cost-cutting measure being deployed by tech firms everywhere.

Meanwhile, in 30 cities with a high volume of technology tenants, office space up for sublease is at its highest levels in a decade, according to CBRE. The volume of new office space leased by tech companies in the fourth quarter was roughly half of the space leased in the same quarter in 2019.

The loss of tech tenants isnt only damaging to office landlords. It can also spell a dip in business for retail landlords in central business districts who rely on steady foot traffic to draw in business.

The vacancy rate of offices across the country rose to 19.8 percent in the first quarter, according to preliminary data from Moodys Analytics, representing a record for its tracker.

Holden Walter-Warner

Read this article:

Tech Firms Retreating From Office Market - The Real Deal

Reset Australia says Meta does not label all untrue content. Photo: Shutterstock

Digital advocacy group Reset Australia has sounded a clear alarm over voluntary big tech regulation and called on the federal government to step in after its complaint about Facebooks handling of misinformation was thrown out.

Reset Australia this week went public with its five-month long battle with Facebook parent company Meta and big tech representative mob Digital Industry Group Inc (DIGI) over the Australian Code of Practice on Disinformation and Misinformation.

Reset Australia has accused DIGI of a conflict of interest between its role in overseeing this code and in receiving funding from many of the companies subject to it.

It has urged the federal government to ditch its practice of using voluntary codes to oversee the tech sector and instead turn to big stick regulation.

In response, DIGI emphasised that the subcommittee handling complaints under the code is independent, that the industry group does not have a vote on it, and that voluntary industry codes designed by industry groups are common practice across a number of sectors.

The issue came to a head as a result of a statement made by Meta in its annual transparency report under the code, with the company saying that it has been applying a warning label to content found to be false by third-party fact-checking organisations.

Reset argued that this statement is untrue as Meta does not actually label all content that contains fact-checked falsehoods.

Metas policy involves the use of AI to find identical or near-identical versions of a post that has been found to be untrue by the independent fact-checkers.

Reset said that despite Meta claiming it is labelling all pieces of misinformation, many posts are slipping through the gap if they are merely worded slightly differently or even posted in a different colour.

Reset Australia produced 152 posts found to be repeating fact-checked falsehoods and reported these to Meta, but just 8 per cent of these posts were labelled.

In response to the letter, Meta agreed to provide further clarification on this process in this years annual report, but refused to make a public correction.

This led Reset to trigger the complaints process under the misinformation code, with an independent subcommittee meeting in March to discuss it.

This week the committee opted to reject Reset Australias complaint, finding that Meta had made a fair and reasonable offer to update its next report with additional information, but that this was rejected.

Reset said this is a mischaracterisation of what occurred.

We requested Meta provide further information so stakeholders could meaningfully verify the claims they made in their last transparency report, Reset Australia said in a statement.

We thanked Meta for their suggestion to add clarifying language in their next transparency report, but urged them to make a public correction. Meta refused to do that, and DIGI has endorsed their position.

Voluntary regulations

The decision reflects the weakness of the voluntary misinformation code, Reset Australia executive director Alice Dawkins said.

The decision today indicates that platforms can say what they like in their transparency reports and their source of independent audit does not have the capacity to run significant data testing of platforms claims, Dawkins said.

Reset Australia is now pushing for the federal government to step in and implement binding regulations and laws to oversee these big tech firms.

The Minister for Communications has said doing nothing is not an option, on misinformation. Well, were sounding a clear alarm that its time for government to do something, Dawkins said.

Big tech have carte blanche in Australia to make whatever decisions they like on our information environment, shroud those systems and processes in PR waffle and block critical scrutiny.

DIGI responds

DIGI managing director Sunita Bose in response said that it is common practice across a number of industries for these codes to be developed by industry associations.

As DIGI is an industry association for the technology sector, we have put strict firewalls in place to avoid conflicts of interest in our administration of the Australian Code of Practice on Disinformation and Misinformation, Bose said.

Bose said that the members of the complaints subcommittee are compensated by DIGI but are contractually independent, and that DIGI is the secretary of the subcommittee but cannot vote on its decisions.

In its decision, the independent subcommittee said that Reset Australia had produced no convincing evidence that Metas transparency report contained false statements, and that Meta has made information on its labelling and fact-checking processes available publicly.

Original post:

Time for government to regulate big tech - ACS

Artificial Intelligence & Machine Learning , Next-Generation Technologies & Secure Development

The British antitrust authority warned Thursday that the market for foundational models in generative artificial intelligence is taking on "winner takes all" dynamics that could entrench a small number of providers.

See Also: 9 Best Practices for Artifact Management

U.K. Competition and Markets Authority head Sarah Cardell in a speech said the regulator won't repeat the hands-off approach that characterized its approach to digital markets over the past decade.

"Could we have responded faster to the competitive threats posed by large digital platforms, recognizing earlier that the tools we had would need adapting to address the unique challenges posed by this new breed of businesses? With the benefit of hindsight, I think the answer is almost certainly 'yes,'" she told a Washington, D.C. audience.

The same patterns of "incumbent firms leveraging their core market power to obstruct new entrants" is again present in the nascent generative AI market, Cardell said.

She pointed to a paper published that day by the CMA concluding that Google, Amazon, Microsoft, Meta and Apple - collectively dubbed the GAMMA firms - are presences up and down the foundational model value chain, raising the specter of greater levels of vertical integration.

Four firms with an outsize presence in offering compute or data resources - Amazon, Google, Meta and Microsoft - could restrict access to critical inputs. Incumbents for consumer or business facing markets could distort choices and restrict competition in foundational model deployment, the report states.

In addition to vertical integration, the regulator sounded a warning over a proliferation of partnerships and strategic investments, identifying an "interconnected web" of more than 90 firms involving GAMMA firms plus chip designer Nvidia

While partnerships could assist independent developers and increase competition, "we are also vigilant against the possibility that incumbent firms may try to use partnerships and investments to quash competitive threats, even where it is uncertain whether those threats will materialize," the report reads.

A Microsoft spokesperson said the company welcomes the "clarity and transparency provided by the report," and that the company is looking forward to engaging constructively with the UK CMA.

The report comes after some British lawmakers warned the competition regulator of potential risk arising from a small number of the largest tech companies influencing the policy developments in the country to use their products to power smaller models for market monopoly (see: UK Market Regulator Reviews Microsoft's Interest in OpenAI ).

Microsoft's close relationship with OpenAI faces similar scrutiny in the European Union. The European competition regulator is additionally investigating $16.3 million, multiyear partnership with Paris-based Mistral AI (see: EU to Analyze Partnership Between Microsoft and Mistral AI).

View post:

UK Markets Authority Warns of AI Market Capture by Big Tech - BankInfoSecurity.com