New Technique Reveals Centuries of Secrets in Locked Letters

The computer-generated unfolding sequence of a sealed letter.Credit...Unlocking History Research Group
The computer-generated unfolding sequence of a sealed letter.Credit...Unlocking History Research Group
TT

New Technique Reveals Centuries of Secrets in Locked Letters

The computer-generated unfolding sequence of a sealed letter.Credit...Unlocking History Research Group
The computer-generated unfolding sequence of a sealed letter.Credit...Unlocking History Research Group

In 1587, hours before her beheading, Mary, Queen of Scots, sent a letter to her brother-in-law Henry III, King of France. But she didn’t just sign it and send it off. She folded the paper repeatedly, cut out a piece of the page, and left it dangling. She used that strand of paper to sew the letter tight with locking stitches.

In an era before sealed envelopes, this technique, now called letterlocking, was as important for deterring snoops as encryption is to your email inbox today. Although this art form faded in the 1830s with the advent of mass-produced envelopes, it has recently attracted renewed attention from scholars. But they have faced a problem: How do you look at the contents of such locked letters without permanently damaging priceless bits of history?

On Tuesday, a team of 11 scientists and scholars at the Massachusetts Institute of Technology and other institutions disclosed their development of a virtual-reality technique that lets them perform this delicate task without tearing up the contents of historical archives.

In the journal Nature Communications, the team tells of virtually opening four undelivered letters written from 1680 and 1706. The dispatches had ended up in a wooden postal trunk in The Hague. Known as the Brienne Collection, the box contains 3,148 items, including 577 letters that were never unlocked.

The new technique could open a window into the long history of communications security. And by unlocking private intimacies, it could aid researchers studying stories concealed in fragile pages found in archives all over the world.

“Let’s start virtually unfolding” the locked letters, said Daniel S. Smith, a team member at King’s College London, “and seeing what secrets they reveal.”

In an interview, Jana Dambrogio, the study’s lead author and a conservator at the M.I.T. Libraries, said that learning of the trove’s existence inspired her to see if more technologically inclined colleagues could find a way to digitally open the locked letters. At the time, in 2014, scholars could read and study such letters only by cutting them open, which often damaged the documents and obscured or eliminated clues as to how they had been secured.

“We really need to keep the originals,” Ms. Dambrogio said. “You can keep learning from them, especially if you keep the locked packets closed.”

The old letters were protected from prying eyes when the sheets of writing paper were carefully folded to become their own secure enclosures.

The first step of their digital opening is to scan a target letter with an advanced X-ray machine. The resulting three-dimensional image — much like a medical scan — reveals the letter’s internal configuration. A computer then analyzes the image to undo the folds and, almost magically, turn the layers into a flat sheet, revealing handwritten text that can be read.

The team translated one of the digitally opened letters from the Brienne Collection. It was dated July 31, 1697, and sent from Lille, France, to a French merchant in The Hague. It turned out to be a request for a certified copy of a death notice. The letter also asked for “news of your health.”

More analyses of the Brienne Collection, the paper added, may enrich studies not only of postal networks in early modern Europe but of the region’s politics, religion, music, drama, and patterns of migration.

In addition to announcing its technique for unlocking the letters without damaging them, the team studied 250,000 historical letters that resulted in “the first systematization of letterlocking techniques.” The scientists and scholars found 12 formats of locked letters — the most complex having an overall shape defined by 12 borders — as well as 64 categories involving such manipulations as tucks, slits, and folds. The team gave each locked letter a security score.

Dr. Smith of King’s College London, who lectures on early modern English literature, said the art was so diverse that a person’s lock could serve almost as a signature. A letter, he said, “became an ambassador for you and had to embody something of you.”

Without the ability to unlock letters digitally, it took a decade for scholars to conclude that Mary, Queen of Scots, had secured the letter to her brother-in-law with a distinctive spiral stitch. Virtual unfolding, the team said, could have documented that step “in a matter of days.”

And Amanda Ghassaei, one of the M.I.T. researchers, said the team was about to complete an upgrade of the computer code that would reduce the time for a virtual unfolding from days to hours.

Deborah Harkness, a historian of science at the University of Southern California who was not involved in the research, described the X-ray technique as “almost an archaeological approach” that seeks to minimize investigator impact on artifact recovery.

The new technique presents “significant innovations,” said Howard Hotson, a professor of early modern intellectual history at the University of Oxford who also had no role in the research. For 500 years, he added, letterlocking, a relatively simple technology, prevented anyone from examining correspondence unless a seal was broken.

“It has taken some very sophisticated digital technology,” he said, “to frustrate this sophisticated security system.”

(The New York Times)



India Eyes $200B in Data Center Investments as It Ramps Up Its AI Hub Ambitions

FILE -Google CEO Sundar Pichai, right, interacts with India's Minister for Information and Technology Ashwini Vaishnaw during Google for India 2022 event in New Delhi, Dec. 19, 2022. (AP Photo/Manish Swarup), File)
FILE -Google CEO Sundar Pichai, right, interacts with India's Minister for Information and Technology Ashwini Vaishnaw during Google for India 2022 event in New Delhi, Dec. 19, 2022. (AP Photo/Manish Swarup), File)
TT

India Eyes $200B in Data Center Investments as It Ramps Up Its AI Hub Ambitions

FILE -Google CEO Sundar Pichai, right, interacts with India's Minister for Information and Technology Ashwini Vaishnaw during Google for India 2022 event in New Delhi, Dec. 19, 2022. (AP Photo/Manish Swarup), File)
FILE -Google CEO Sundar Pichai, right, interacts with India's Minister for Information and Technology Ashwini Vaishnaw during Google for India 2022 event in New Delhi, Dec. 19, 2022. (AP Photo/Manish Swarup), File)

India is hoping to garner as much as $200 billion in investments for data centers over the next few years as it scales up its ambitions to become a hub for artificial intelligence, the country’s minister for electronics and information technology said Tuesday.

The investments underscore the reliance of tech titans on India as a key technology and talent base in the global race for AI dominance. For New Delhi, they bring in high-value infrastructure and foreign capital at a scale that can accelerate its digital transformation ambitions.

The push comes as governments worldwide race to harness AI's economic potential while grappling with job disruption, regulation and the growing concentration of computing power in a few rich countries and companies.

“Today, India is being seen as a trusted AI partner to the Global South nations seeking open, affordable and development-focused solutions,” Ashwini Vaishnaw told The Associated Press in an email interview, as New Delhi hosts a major AI Impact Summit this week drawing participation from at least 20 global leaders and a who’s who of the tech industry.

In October, Google announced a $15 billion investment plan in India over the next five years to establish its first artificial intelligence hub in the South Asian country. Microsoft followed two months later with its biggest-ever Asia investment announcement of $17.5 billion to advance India’s cloud and artificial intelligence infrastructure over the next four years.

Amazon too has committed $35 billion investment in India by 2030 to expand its business, specifically targeting AI-driven digitization. The cumulative investments are part of $200 billion in investments that are in the pipeline and New Delhi hopes would flow in.

Vaishnaw said India’s pitch is that artificial intelligence must deliver measurable impacts at scale rather than remain an elite technology.

“A trusted AI ecosystem will attract investment and accelerate adoption,” he said, adding that a central pillar of India’s strategy to capitalize on the use of AI is building infrastructure.

The government recently announced a long-term tax holiday for data centers as it hopes to provide policy certainty and attract global capital.

Vaishnaw said the government has already operationalized a shared computing facility with more than 38,000 graphics processing units, or GPUs, allowing startups, researchers and public institutions to access high-end computing without heavy upfront costs.

“AI must not become exclusive. It must remain widely accessible,” he said.

Alongside the infrastructure drive, India is backing the development of sovereign foundational AI models trained on Indian languages and local contexts. Some of these models meet global benchmarks and in certain tasks rival widely used large language models, Vaishnaw said.

India is also seeking a larger role in shaping how AI is built and deployed globally as the country doesn’t see itself strictly as a “rule maker or rule taker,” according to Vaishnaw, but an active participant in setting practical, workable norms while expanding its AI services footprint worldwide.

“India will become a major provider of AI services in the near future,” he said, describing a strategy that is “self-reliant yet globally integrated” across applications, models, chips, infrastructure and energy.

Investor confidence is another focus area for New Delhi as global tech funding becomes more cautious.

Vaishnaw said the technology’s push is backed by execution, pointing to the Indian government's AI Mission program which emphasizes sector specific solutions through public-private partnerships.

The government is also betting on reskilling its workforce as global concerns grow that AI could disrupt white collar and technology jobs. New Delhi is scaling AI education across universities, skilling programs and online platforms to build a large AI-ready talent pool, the minister said.

Widespread 5G connectivity across the country and a young, tech-savvy population are expected to help with the adoption of AI at a faster pace, he added.

Balancing innovation with safeguards remains a challenge though, as AI expands into sensitive sectors such as governance, health care and finance.

Vaishnaw outlined a fourfold strategy that includes implementable global frameworks, trusted AI infrastructure, regulation of harmful misinformation and stronger human and technical capacity to hedge the impact.

“The future of AI should be inclusive, distributed and development-focused,” he said.


Report: SpaceX Competing to Produce Autonomous Drone Tech for Pentagon 

The SpaceX logo is seen in this illustration taken, March 10, 2025. (Reuters)
The SpaceX logo is seen in this illustration taken, March 10, 2025. (Reuters)
TT

Report: SpaceX Competing to Produce Autonomous Drone Tech for Pentagon 

The SpaceX logo is seen in this illustration taken, March 10, 2025. (Reuters)
The SpaceX logo is seen in this illustration taken, March 10, 2025. (Reuters)

Elon Musk's SpaceX and its wholly-owned subsidiary xAI are competing in a secret new Pentagon contest to produce voice-controlled, autonomous drone swarming technology, Bloomberg News reported on Monday, citing people familiar with the matter.

SpaceX, xAI and the Pentagon's defense innovation unit did not immediately respond to requests for comment. Reuters could not independently verify the report.

Texas-based SpaceX recently acquired xAI in a deal that combined Musk's major space and defense contractor with the billionaire entrepreneur's artificial intelligence startup. It occurred ahead of SpaceX's planned initial public offering this year.

Musk's companies are reportedly among a select few chosen to participate in the $100 million prize challenge initiated in January, according to the Bloomberg report.

The six-month competition aims to produce advanced swarming technology that can translate voice commands into digital instructions and run multiple drones, the report said.

Musk was among a group of AI and robotics researchers who wrote an open letter in 2015 that advocated a global ban on “offensive autonomous weapons,” arguing against making “new tools for killing people.”

The US also has been seeking safe and cost-effective ways to neutralize drones, particularly around airports and large sporting events - a concern that has become more urgent ahead of the FIFA World Cup and America250 anniversary celebrations this summer.

The US military, along with its allies, is now racing to deploy the so-called “loyal wingman” drones, an AI-powered aircraft designed to integrate with manned aircraft and anti-drone systems to neutralize enemy drones.

In June 2025, US President Donald Trump issued the Executive Order (EO) “Unleashing American Drone Dominance” which accelerated the development and commercialization of drone and AI technologies.


SVC Develops AI Intelligence Platform to Strengthen Private Capital Ecosystem

The platform offers customizable analytical dashboards that deliver frequent updates and predictive insights- SPA
The platform offers customizable analytical dashboards that deliver frequent updates and predictive insights- SPA
TT

SVC Develops AI Intelligence Platform to Strengthen Private Capital Ecosystem

The platform offers customizable analytical dashboards that deliver frequent updates and predictive insights- SPA
The platform offers customizable analytical dashboards that deliver frequent updates and predictive insights- SPA

Saudi Venture Capital Company (SVC) announced the launch of its proprietary intelligence platform, Aian, developed in-house using Saudi national expertise to enhance its institutional role in developing the Kingdom’s private capital ecosystem and supporting its mandate as a market maker guided by data-driven growth principles.

According to a press release issued by the SVC today, Aian is a custom-built AI-powered market intelligence capability that transforms SVC’s accumulated institutional expertise and detailed private market data into structured, actionable insights on market dynamics, sector evolution, and capital formation. The platform converts institutional memory into compounding intelligence, enabling decisions that integrate both current market signals and long-term historical trends, SPA reported.

Deputy CEO and Chief Investment Officer Nora Alsarhan stated that as Saudi Arabia’s private capital market expands, clarity, transparency, and data integrity become as critical as capital itself. She noted that Aian represents a new layer of national market infrastructure, strengthening institutional confidence, enabling evidence-based decision-making, and supporting sustainable growth.

By transforming data into actionable intelligence, she said, the platform reinforces the Kingdom’s position as a leading regional private capital hub under Vision 2030.

She added that market making extends beyond capital deployment to shaping the conditions under which capital flows efficiently, emphasizing that the next phase of market development will be driven by intelligence and analytical insight alongside investment.

Through Aian, SVC is building the knowledge backbone of Saudi Arabia’s private capital ecosystem, enabling clearer visibility, greater precision in decision-making, and capital formation guided by insight rather than assumption.

Chief Strategy Officer Athary Almubarak said that in private capital markets, access to reliable insight increasingly represents the primary constraint, particularly in emerging and fast-scaling markets where disclosures vary and institutional knowledge is fragmented.

She explained that for development-focused investment institutions, inconsistent data presents a structural challenge that directly impacts capital allocation efficiency and the ability to crowd in private investment at scale.

She noted that SVC was established to address such market frictions and that, as a government-backed investor with an explicit market-making mandate, its role extends beyond financing to building the enabling environment in which private capital can grow sustainably.

By integrating SVC’s proprietary portfolio data with selected external market sources, Aian enables continuous consolidation and validation of market activity, producing a dynamic representation of capital deployment over time rather than relying solely on static reporting.

The platform offers customizable analytical dashboards that deliver frequent updates and predictive insights, enabling SVC to identify priority market gaps, recalibrate capital allocation, design targeted ecosystem interventions, and anchor policy dialogue in evidence.

The release added that Aian also features predictive analytics capabilities that anticipate upcoming funding activity, including projected investment rounds and estimated ticket sizes. In addition, it incorporates institutional benchmarking tools that enable structured comparisons across peers, sectors, and interventions, supporting more precise, data-driven ecosystem development.