Amazon.com Inc. reported hundreds of thousands of pieces of content last year that it believed included child sexual abuse, which it found in data gathered to improve its artificial intelligence models. Though Amazon removed the content before training its models, child safety officials said the company has not provided information about its source, potentially hindering law enforcement from finding perpetrators and protecting victims. 

Throughout last year, Amazon detected the material in its AI training data and reported it to the National Center for Missing and Exploited Children, or NCMEC. The organization, which was established by Congress to field tips about child sexual abuse and share them with law enforcement, recently started tracking the number of reports specifically tied to AI products and their development. In 2025, NCMEC saw at least a fifteen-fold increase in these AI-related reports, with “the vast majority” coming from Amazon. The findings haven’t been previously reported.

An Amazon spokesperson said the training data was obtained from external sources, and the company doesn’t have the details about its origin that could aid investigators. It’s common for companies to use data scraped from publicly available sources, such as the open web, to train their AI models. Other large tech companies have also scanned their training data and reported potentially exploitative material to NCMEC. However, the clearinghouse pointed to “glaring differences” between Amazon and its peers. The other companies collectively made just “a handful of reports,” and provided more detail on the origin of the material, a top NCMEC official said.

In an emailed statement, the Amazon spokesperson said that the company is committed to preventing child sexual abuse material across all of its businesses. “We take a deliberately cautious approach to scanning foundation model training data, including data from the public web, to identify and remove known [child sexual abuse material] and protect our customers,” the spokesperson said. 

The spike in Amazon’s reports coincides with a fast moving AI race that has left companies large and small scrambling to acquire and ingest huge volumes of data to improve their models. But that race has also complicated the work of child safety officials — who are struggling to keep up with the changing technology — and challenged regulators tasked with safeguarding AI from abuse. AI safety experts warn that quickly amassing large datasets without proper safeguards comes with grave risks. 

Amazon accounted for most of the more than 1 million AI-related reports of child sexual abuse material submitted to NCMEC in 2025, the organization said. It marks a jump from the 67,000 AI-related reports that came from across the tech and media industry a year prior, and just 4,700 in 2023. This category of AI-related reports can include AI-generated photos and videos, or sexually explicit conversations with AI chatbots. It can also include photos of real victims of sexual abuse that were collected, even unintentionally, in an effort to improve AI models. 

Training AI on illegal and exploitative content raises newfound concerns. It could risk shaping a model’s underlying behaviors, potentially improving its ability to digitally alter and sexualize photos of real children or create entirely new images of sexualized children that never existed. It also raises the threat of continuing the circulation of the images that models were trained on — re-victimizing children who have suffered abuse. 

The Amazon spokesperson said that, as of January, the company is “not aware of any instances” of its models generating child sexual abuse material.  None of its reports submitted to NCMEC were of AI-generated material, the spokesperson added. Instead, the content was flagged by an automatic detection tool that compared it against a database of known child abuse material involving real victims, a process called “hashing.” Approximately 99.97% of the reports resulted from scanning “non-proprietary training data,” the spokesperson said. 

Amazon believes it over-reported these cases to NCMEC to avoid accidentally missing something. “We intentionally use an over-inclusive threshold for scanning, which yields a high percentage of false positives,” the spokesperson added.

Amazon has more than 900 data center facilities worldwide.
Photo Credit: Bloomberg

The AI-related reports received last year are just a fraction of the total number submitted to NCMEC. The larger category of reports also includes suspected child sexual abuse material sent in private messages or uploaded to social media feeds and the cloud. In 2024, for example, NCMEC received more than 20 million reports from across industry, with most coming from Meta Platforms Inc. subsidiaries Facebook, Instagram and WhatsApp. Not all reports are ultimately confirmed as containing child sexual abuse material, referred to with the acronym CSAM.

Still, the volume of suspected CSAM that Amazon detected across its AI pipeline in 2025 stunned child safety experts interviewed by Bloomberg News. The hundreds of thousands of reports made to NCMEC marked a drastic surge for the company. In 2024, Amazon and all of its subsidiaries made a total of 64,195 reports. 

“This is really an outlier,” said Fallon McNulty, the executive director of NCMEC’s CyberTipline, the entity to which US-based social media platforms, cloud providers and other companies are legally required to report suspected CSAM. “Having such a high volume come in throughout the year begs a lot of questions about where the data is coming from, and what safeguards have been put in place.” 

McNulty, speaking in an interview, said she has little visibility into what’s driving the surge of sexually exploitative material in Amazon’s initial training data sets. Amazon has provided “very little to almost no information” in their reports about where the illicit material originally came from, who had shared it, or if it remains actively available on the internet, she said. 

While Amazon is not required to share this level of detail, the lack of information makes it impossible for NCMEC to track down the material’s origin and work to get it removed, McNulty said. It also limits relevant law enforcement agencies tasked with searching for sex offenders and children in active danger. “There’s nothing then that can be done with those reports,” she said. “Our team has been really clear with [Amazon] that those reports are inactionable.”

When asked why the company didn’t disclose information about the possible origin of the material, or other key details, the Amazon spokesperson replied, “because of how this data is sourced, we don’t have the data that comprises an actionable report.” The spokesperson did not explain how the third-party data was sourced or why the company did not have sufficient information to create actionable reports. “While our proactive safeguards cannot provide the same detail in NCMEC reports as consumer-facing tools, we stand by our commitment to responsible AI and will continue our work to prevent CSAM,” the spokesperson said.

NCMEC, a nonprofit, receives funding both from the US government and private industry. Amazon is among its funders and holds a corporate seat on its board. 

“There should be more transparency on how companies are gathering and analyzing the data to train their models — and how they’re training them,” said David Thiel, the former chief technologist at the Stanford Internet Observatory, who has researched the prevalence of child sexual abuse material in AI training data. 

Such data can be licensed, purchased or scraped from the internet, or could be so-called synthetic data, which is text or images created by other AI tools. As AI companies seek to release new models quickly, “the rapid gathering of data is a much higher priority than doing safety analyses,” Thiel said. He warned that there are “always some errors” when it comes to sifting out CSAM from training data, and believes the industry needs to be more open about where its data is coming from. 

Amazon’s Bedrock offering, which gives customers access to various AI models so they can build their own AI products, includes automated detection for known CSAM and rejects and reports positive matches. The company’s consumer-facing generative AI products also allow users to report content that escapes its controls. 

The Seattle-based tech giant scans for CSAM across its other businesses, too, including its consumer photo storage service. Amazon’s cloud computing division, Amazon Web Services, also removes CSAM when it’s discovered on the web services it hosts. McNulty said AWS submitted far fewer reports than came from Amazon’s AI efforts. Amazon declined to break out specific reporting data across its various business units, but noted it would share broad data in March.

Only recently have technology companies really begun to scrutinize their AI models and training data for CSAM, said David Rust-Smith, a data scientist at Thorn, a nonprofit organization that provides tools to companies, including Amazon, to detect the exploitative material. 

“There’s definitely been a big shift in the last year of people coming to us asking for help cleaning data sets,” said Rust-Smith. He noted that “some of the biggest players” have sought to apply Thorn’s detection tools to their training data, but declined to speak about any individual company. Amazon did not use Thorn’s technology to scan its training data, the spokesperson confirmed.

Rust-Smith said AI-focused companies are approaching Thorn with a newfound urgency. “People are learning what we already knew, which is, if you hoover up a ton of the internet, you’re going to get [child sexual abuse material],” he said. 

Amazon was not the only company to spot and report potential CSAM from its AI workflows last year. Alphabet Inc.’s Google and OpenAI told Bloomberg News that they scan AI training data for exploitative material — a process that has surfaced potential CSAM, which the  companies then reported to NCMEC. Meta and Anthropic PBC said they, too, search training data for CSAM. Meta did not comment on whether it had identified the material, but said it would report to NCMEC if it did. Anthropic said it has not reported such material out of its training data. Meta and Google said that they’ve taken efforts to ensure that reports related to their AI workflows are distinguishable from those generated by others parts of their business. 

McNulty said that, with the exception of Amazon, the AI-related reports it received last year came in “really, really small volumes,” and included key details that allowed the clearinghouse to pass on actionable information to law enforcement. 

“Simply flagging that you came across something but not providing any type of actionable detail doesn’t help the larger child safety space,” McNulty said.

Tata Consultancy Services (TCS), Cognizant and eight other major corporations have been questioned by U.S. Senators for filing thousands of H-1B skilled labour visa petitions after conducting “mass layoffs” of American employees.

Senate Judiciary Committee Chairman Charles Grassley and ranking member Richard Durbin have also questioned Amazon, Apple, Deloitte, Google, JPMorgan Chase, Meta, Microsoft, and Walmart, asking them for details on their hiring and recruitment practices as well as any variation in salary and benefits between H-1B visa holders and American employees.

The Senators said their “inquiry” comes at a time when the unemployment rate in America’s tech sector is “well above” the overall jobless rate. According to the Federal Reserve, recent American graduates with STEM degrees now face higher unemployment rates than the general population, they stated.

It is pertinent to mention that Senator Grassley and Senator Durbin have been vocal critics of the H-1B visa programme and have consistently contended that these visas are being abused by many to replace U.S. workers with cheaper labour from overseas.

In fact, just a few days after their missive to the ten companies, the Senators said they are reintroducing bipartisan legislation to “reform and close loopholes” in the H-1B and L-1 visa programs.

The H-1B and L-1 Visa Reform Act targets fraud and abuse in U.S. immigration system, provides protection for American workers and visa holders and enhances transparency in the foreign worker recruitment process, the two argued.

“In evaluating the high unemployment rate for American tech workers, we cannot ignore the massive, ongoing layoffs ordered by you and your peers in Big Tech C-suites over the past few years…At the same time you have been laying off your employees, you have been filing H-1B visa petitions for (thousands of) foreign workers,” the Senators wrote in letters to 10 major employers in the U.S. on September 25, 2025.

They added: “With all of the homegrown American talent relegated to the sidelines, we find it hard to believe that (you) cannot find qualified American tech workers to fill these positions.”

On TCS, the statement noted that the company recently announced plans to lay off over 12,000 employees, including American staff.

“In fiscal year 2025, TCS applied for and received approval to hire 5,505 foreign H1-1B employees. TCS is currently under investigation by the Equal Employment Opportunity Commission for allegedly firing older American employees in favour of newly hired foreign H-1B visa holders,” it said.

It further said Cognizant Technology Solutions laid off thousands of workers, including employees in the US.

“In fiscal year 2025, Cognizant applied for and received approval to hire 2,493 foreign H-1B employees. In 2024, a federal jury concluded Cognizant favoured South Asian H1-B visa holders over American employees, which warranted punitive damages,” it added.

E-mails sent to TCS and Cognizant for comments by PTI did not elicit a response.

The statement by Grassley and Durbin noted that Amazon laid off tens-of-thousands of employees in recent years, blaming the adoption of generative AI tools. In fiscal year 2025, Amazon applied for and received approval to hire 10,044 foreign H-1B employees.

Google, it said, laid off tens-of-thousands of employees in recent years, including hundreds of workers in its platform and device unit, and 35% of its small team managers “despite enjoying records profits”.

In fiscal year 2025, Google applied for and received approval to hire 4,181 foreign H-1B employees, the statement observed.

According to it, Meta laid off a quarter of its workforce between 2022 and 2023 and labelled it a “year of efficiency”.

“In 2025, Meta laid off 3,600 employees, overshooting its stated goal of laying off another five per cent of its workforce. In fiscal year 2025, Meta applied for and received approval to hire 5,123 foreign H-1B employees,” it said.

Microsoft applied for and received approval to hire 5,189 foreign H-1B employees in fiscal year 2025, as per the statement.

Microsoft laid off 16,000 employees this year alone, despite record revenue and profits, it said adding amid criticism for mass-layoffs, Microsoft characterised its actions as an “enigma of success”.

Published – October 03, 2025 10:49 pm IST

An Indian startup founder’s story of how he went from generating a revenue of Rs 20 lakh per day to almost losing everything is going viral on social media. Shared by Grapevine founder Saumil Tripathi on X (formerly Twitter), the story shows how e-commerce giant Amazon capitalised on the founder’s business model and managed to scale it, eventually driving him out of the marketplace.

“I went from selling 20L of products per day to watching my generational wealth dream crumble”. An e-commerce founder shared the story of their rise and fall on Amazon! [sic],” read the post by Mr Tripathi on X.

As per the unnamed founder, he started a home organiser company in 2017 which quickly began to provide incredible returns.

“I was on AliExpress, looking for budget-friendly storage ideas for my own apartment-think suction-cup shelves, collapsible bins, drawers,” the entrepreneur shared in his post, adding that he realised that these products were way more expensive on Amazon.

After spending Rs 2.5 lakh initially to buy Chinese products and quickly reselling them in India, the entrepreneur expanded the business by stockpiling the inventory. He even went to Chinese factories to directly source the product which increased his profit margins.

The business was booming and soon Amazon approached the founder and offered him a “nine-figure” buyout. “They pitched me a collaboration or potential acquisition, hinting that my brand complemented their private label push,” the post read further.

However, the startup declined the offer and it was to be the beginning of the end. Seeing the success of the products, Amazon introduced its own private-label brand, Solimo. With eerily similar products but at significantly lower prices, Solimo began to dominate search results on Amazon, pushing the entrepreneur’s offerings off the virtual shelves.

The OP’s top-selling items, once leaders in their category, found themselves overshadowed by Amazon’s strategic move. If he tried to reduce the prices, Amazon slashed the margins further until the OP was forced to clear the inventory on loss.

“Today, that business is practically gone, undone by Amazon’s move into private labels. I’m not broke or working a 9-5, but the potential for creating true generational wealth was ripped out from under me before it could fully materialise. This is my cautionary tale,” he said.

“I went from selling 20L of products per day to watching my generational-wealth dream crumble”

An e-commerce founder shared the story of their rise and fall on Amazon! pic.twitter.com/jvZl5PNDus

— Saumil Heard It (@OnTheGrapevine) December 26, 2024

Internet reacts

As of the last update, the post had garnered 1.2 million views, nearly 7,000 likes and hundreds of comments. Most of the users expressed sympathy with the founder but others added that running a business without a unique product was always supposed to end in pain.

“This might be a tough read. But it is Biz101,” said one user, while another added: “It’s a leverage game and marketplaces have the ultimate leverage.”

A third commented: “Please don’t think that Amazon is your business partner. You spend on ads and they get your customer’s data. Once you grow big they will use your customer data and start selling themselves.

The founder signed off by cautioning up-and-coming entrepreneurs to be alert when an acquisition offer comes and also advised that developing a unique product was essential to surviving on such platforms.

New Delhi:

E-commerce giant Amazon on Wednesday announced that Samir Kumar will take on the responsibility to lead India as its country manager, following the resignation of Manish Tiwary.

According to a leadership update by the company, Mr Kumar is working closely with Manish Tiwary on the transition, and will assume operational responsibilities for India starting October 1.

“Amazon’s 25-year veteran Samir Kumar will oversee Amazon’s India consumer business as Manish Tiwary, current Country Manager for Amazon India, has decided to explore an opportunity outside Amazon,” the update said.

Amit Agarwal, Amazon’s senior Vice President for Emerging Markets, in an internal email, announced that Kumar will take on the responsibility to lead India as its Country Manager, in addition to his current charter of leading consumer businesses in the Middle East, South Africa and Turkey.

Mr Kumar, who joined Amazon in 1999, was a part of the original team that planned and launched Amazon.in in 2013.

With this change, the current Amazon.in leadership team of Saurabh Srivastava (Categories), Harsh Goyal (Everyday Essentials), Amit Nanda (Marketplace), and Aastha Jain (Growth Initiatives) will now report to Kumar.

Kishore Thota (Emerging Markets Shopping Experience) will report to Mr Agarwal directly.

“India remains an important priority for Amazon, and I am super excited about the opportunity ahead as we continue to transform lives and livelihoods. We have a strong local leadership bench and, along with Samir’s experiences across Emerging Markets, I am even more optimistic about our future plans to deliver for customers and the business in India,” Agarwal said.

Mr Agarwal also expressed appreciation for Mr Tiwary’s leadership “in steering Amazon.in to become the de facto starting point for Indians to buy and sell anything online” and wished him the best for future endeavours. 

Amazon.com said on Monday it would require employees to return to the office five days per week, effective January 2, 2025.

“We are also going to bring back assigned desk arrangements in locations that were previously organized that way, including the U.S. headquarters locations (Puget Sound and Arlington),” CEO Andy Jassy said in a note to employees.

As part of an organizational restructuring, Amazon is looking to increase the ratio of individual contributors to managers by at least 15% by the end of the first quarter of 2025.

In May last year, employees at Amazon’s Seattle headquarters staged a walkout protesting changes to the ecommerce giant’s climate policy, layoffs and a return-to-office mandate.

(Except for the headline, this story has not been edited by NDTV staff and is published from a syndicated feed.)

A report by Senator Bernie Sanders’ office found Amazon failed to adequately staff its warehouses during peak shopping times.
| Photo Credit: AP

Amazon’s popular Prime Day sales event has been “a major cause of injuries” for warehouse workers who pick and pack customer orders at the e-commerce giant’s facilities across the United States, according to a report released Tuesday by Senator Bernie Sanders.

The report, which draws information from a year-long Senate committee investigation into Amazon’s safety practices and relied on internal company data from 2019 and 2020, found peak shopping times — including the holiday shopping period — resulted in the “highest weekly injury rates” for warehouse workers.

The preliminary report from Mr. Sanders’ office was also based on interviews with more than 100 current and former Amazon employees. This year’s two-day Prime Day event started Tuesday.

In a statement, Mr. Sanders said the “incredibly dangerous working conditions at Amazon” highlighted in the report are a “perfect example of the type of corporate greed that the American people are sick and tired of”.

“Despite making $36 billion in profits last year and providing its CEO with over $275 million in compensation over the past three years, Amazon continues to treat its workers as disposable and with complete contempt for their safety and well-being,” said the Vermont independent, who has been critical of Amazon and supports worker efforts to unionise at the company. “That is unacceptable, and that has got to change,” he added.

Labour unions and safety experts have long criticised Amazon, alleging the company’s focus on speed and fast deliveries puts workers in danger. In recent years, some States have passed laws aimed at Amazon to curb the use of warehouse productivity quotas, but Amazon claims it doesn’t employ them.

According to the Senate report, 45 out of 100 warehouse workers at Amazon received injuries during the 2019 Prime Day event. The number included minor injuries the company was not required to disclose to the federal government, such as bruises and superficial cuts, but also serious ones such as concussions that should have been reported, it said.

Amazon denies finding

Amazon disputed the finding.

“The claims that we systemically underreport injuries, and that our actual injury rates are higher than publicly reported, are false,” Amazon spokesperson Kelly Nantel said in a prepared statement. “We’re required to report every injury that needs more than basic first aid, and that’s what we do.”

While Amazon “might make an occasional clerical error,” a six-month federal investigation by the Occupational Safety and Health Administration (OSHA) found “no intentional, willful, or systemic errors” in the company’s reporting, Ms. Nantel said.

The report also alleged that Amazon had a practice of failing to refer workers for outside medical care because doing so could affect whether an injury should be considered “recordable” and referred to OSHA. Even when injuries were serious, workers often received first aid before being sent back to work instead of to a doctor, according to the report.

Amazon has acknowledged in the past that its warehouse injury rates had been higher compared to its peers. Federal safety investigators levied fines against the company in recent years following inspections at some of its warehouses. Some of the inspections arose from referrals made to the U.S. Attorney’s Office for the Southern District of New York, which is also investigating worker safety at the company through its civil division.

The report also says Amazon failed to adequately staff its warehouses during peak shopping times, which the company disputed. Amazon said in March that it allocated over $750 million to safety efforts for this year.

A New York federal judge has ordered Amazon.com to comply with a subpoena from a U.S. civil rights agency investigating claims that the online retailer discriminated against pregnant warehouse workers.

U.S. District Judge Lorna Schofield in Manhattan late Thursday rejected Amazon’s claims that the Equal Employment Opportunity Commission (EEOC) subpoena was too broad and sought irrelevant information.

The EEOC is seeking data on requests that pregnant workers at five U.S. warehouses made for accommodations such as limits on heavy lifting and additional breaks, and whether Amazon granted or denied them.

The commission’s probe was prompted by complaints from five women who say they faced pregnancy discrimination while working at Amazon warehouses in New Jersey, Connecticut, North Carolina, and California.

Amazon provided the EEOC with more than 200,000 pages of data in response to the subpoena, but not the specific information requested by the agency, according to court filings.

Schofield in her ruling said the information sought in the subpoenas was necessary for the EEOC to determine whether Amazon engaged in illegal discrimination. The judge gave Amazon until Aug. 9 to comply with the subpoena.

Amazon did not immediately respond to a request for comment on Friday. An EEOC spokesman declined to comment.

In 2022, a New York state agency filed an administrative complaint accusing Amazon of requiring pregnant and disabled warehouse workers to take unpaid leaves of absence, even if they were capable of working, instead of providing accommodations. That case is pending.

Amazon has denied wrongdoing and said it strives to support its workers, but acknowledged in a statement responding to the New York complaint that “we don’t always get it right.”

The EEOC launched its probe last year and issued a subpoena seeking five categories of information, including data on accommodations Amazon provided to warehouse workers with disabilities. At the time, federal law only required companies to provide the same accommodations to pregnant workers that they gave to employees with disabilities.

A law passed later last year mandates that employers accommodate workers’ pregnancies regardless of how they treat workers with disabilities.

(Except for the headline, this story has not been edited by NDTV staff and is published from a syndicated feed.)

Amazon founder and executive chair Jeff Bezos is planning to sell almost $5 billion worth of shares in the e-commerce giant, a regulatory filing showed, after its stock hit a record high.

The proposed sale of 25 million shares was disclosed in a notice filed after market hours on Tuesday. The stock had hit an all-time high of $200.43 during the session. It has jumped more than 30% so far this year, outpacing the 4% gain in the Dow Jones Industrial Average index.

After the sale plan, Bezos would own about 912 million Amazon shares, or 8.8% of the outstanding stock.

He sold shares worth roughly $8.5 billion in February, after the stock rallied 80% in 2023.

Bezos is ranked the second-richest person in the world with a net worth of $214.4 billion, according to Forbes. He is also the founder of space company Blue Origin, which launched a six-person crew to the edge of space in May.

Amazon posted upbeat first-quarter results in April, as the Seattle-based technology giant rode the artificial intelligence wave. The company recently replaced Adam Selipsky as the head of its cloud computing unit with insider Matt Garman.

(Except for the headline, this story has not been edited by NDTV staff and is published from a syndicated feed.)

Amazon.com plans to launch a section on its shopping site featuring cheap items that ship directly to overseas consumers from warehouses in China, the Information reported on Wednesday, citing slides shown to Chinese sellers.

The new marketplace, Amazon’s most aggressive response to the growth of bargain sites like Temu and Shein, will offer unbranded fashion, home goods, and daily necessities, according to the slides, and will be delivered between 9 to 11 days to customers, the report said.

The e-commerce giant in a recent closed-door meeting told Chinese sellers it would start signing up merchants this summer and begin accepting inventory in the fall, according to the Information.

Amazon did not immediately respond to Reuters request for comment on the report.

Sellers joining the bargain site can determine their product selection and pricing, and can produce in small batches to test the demand for any new products they plan to launch, the report added.

It is not clear if these shipments will be made using a U.S. trade provision that exempts individual packages worth less than $800 from US customs duties, Information reported.

E-commerce powerhouse Shein, which is trying to expand its market share before going public, and PDD Group-owned e-retailer Temu, depend on the expedited clearance process, which is available for direct-to-consumer shipments valued at $800 or less.

(Except for the headline, this story has not been edited by NDTV staff and is published from a syndicated feed.)

Milan:

Italy’s antitrust authority has fined two subsidiaries of Amazon 10 million euros ($10.70 million) for alleged unfair commercial practices, the regulator said on Wednesday.

According to the watchdog, Amazon has “significantly restricted consumers freedom of choice” by automatically pre-setting the ‘Subscribe&Save’ option, for regular scheduled deliveries rather than one-off purchases, for a wide selection of products.

“The pre-ticking of recurring purchase induces one to periodically buy a product – even without the effective need – thus limiting one’s freedom of choice,” the antitrust said in a statement.

Amazon was not immediately available for comment.

(Except for the headline, this story has not been edited by NDTV staff and is published from a syndicated feed.)