FCA Crypto Authorisation Gateway Opens as UK Brings Digital Assets Into Full Regulation

FCA Crypto Authorisation Gateway Opens as UK Brings Digital Assets Into Full Regulation

by | Oct 14, 2026 | How to Crypto

The Financial Conduct Authority opened its FCA crypto authorisation gateway on 30 September 2026, allowing firms to apply for permission to carry on newly regulated cryptoasset activities in the UK. The move brings crypto businesses operating in the country into full FCA regulation for the first time, with standards covering consumer protection, safeguarding of customer assets, market integrity and financial resilience.

The gateway was created under the Financial Services and Markets Act 2000 (Cryptoassets) Regulations 2026, which amend the Regulated Activities Order to establish a new set of regulated activities directed at cryptoassets. The regime replicates for crypto businesses the regulation of familiar functions in traditional financial markets, including issuance, custody, trading venues, dealing, intermediation and certain staking services.

Deadlines and Transitional Arrangements

Businesses planning to keep serving the UK market must submit their applications through the FCA crypto authorisation gateway no later than 28 February 2027, with the full regime taking effect on 25 October 2027. The regulator has urged firms to file well ahead of that date instead of treating it as a target, according to JD Supra.

Submitting an application within the prescribed period allows an eligible existing business to benefit from statutory transitional provisions, continuing to provide cryptoasset services, including taking on new business, while the FCA reviews its application after the new regime commences. A firm does not need to have received final authorisation by the February deadline, but those applying after it may struggle to gain approval in time and could need to cease the relevant regulated activities until permission is obtained, JD Supra reported.

Decisions on applications filed within the window are expected ahead of the regime's start date, the regulator said. Applicants must show clearly that they satisfy the regulator's standards, and any firm unable to demonstrate this will be denied authorisation for the UK market.

How to Apply Through FCA Connect

Applications for Part 4A permission covering the new cryptoasset regulated activities can now be submitted via FCA Connect. Firms that already hold FSMA authorisation have the option of seeking a variation of permission instead, adding the relevant cryptoasset activities to what they are currently permitted to do.

A firm's existing regulatory status will not carry over into the new framework on its own. Registration with the FCA under the Money Laundering Regulations does not entitle a crypto business to an automatic Part 4A permission for the corresponding activity from October 2027, according to JD Supra. Businesses with current FSMA authorisation, or those supervised under the Payment Services Regulations or Electronic Money Regulations, must assess whether the permissions they hold extend to their cryptoasset work and seek a variation where they fall short.

The FCA is supporting firms to prepare for authorisation through pre-application discussions and webinars, which are available on demand. Its Pre-Application Support Service remains available, though the regulator expects meeting requests to come with substantive detail rather than promises to supply information at a later stage, JD Supra reported.

What the New Regime Requires

Crypto businesses caught by the rules will face oversight much like that applied to conventional FCA-authorised financial services firms. Obligations will span capital and prudential rules, standards for how business is conducted, governance, systems and controls, the Consumer Duty where it applies, plus regulatory reporting and ongoing supervision.

Senior staff at authorised firms will fall under the regulator's individual accountability framework, which encompasses the senior managers and certification regime. Governance structures must be proportionate to the nature, scale and complexity of each firm's activities, with clear allocation of senior-management responsibilities.

Issuers of stablecoins will face rules on how backing assets are safeguarded, under a new Client Assets Sourcebook (CASS) 16 chapter of the FCA Handbook. Custodians will face specific requirements concerning the safeguarding of client cryptoassets, and firms that outsource custody to third parties must weigh how those arrangements affect their own regulatory obligations.

Once authorised, firms come under continuing FCA oversight that includes reporting duties and information requests, alongside an expectation of open and cooperative dealings with the regulator. The FCA's existing enforcement toolkit will be available in relation to the new regulated activities, including financial penalties, public censures and restrictions or prohibitions where statutory conditions are satisfied.

Scope Reaches Overseas and Decentralised Businesses

A broad set of business models could be captured, spanning exchanges and trading venues, brokers and other intermediaries, custodians and wallet providers, stablecoin issuers, some market makers and principal dealers, and firms that arrange staking services. The perimeter analysis remains highly fact specific, according to JD Supra.

A firm whose principal place of business sits outside the UK may still fall in scope where it has UK employees or agents undertaking regulated dealing or arranging functions, or otherwise operates through substantive UK-based operations. That could include UK servers, UK-targeted marketing efforts or bank accounts. Regulation may also apply where relevant transactions are negotiated, arranged or executed in the UK or through a UK establishment, or where safeguarding arrangements involve the exercise of relevant control from the UK.

Technology providers, introducers, non-custodial businesses and decentralised protocols will need to consider carefully whether their particular activities satisfy the statutory definitions and territorial requirements, and whether an exclusion applies. Exclusions and exemptions will be considered on a fact-specific, case-by-case basis.

Preparing for Authorisation

Dominic Cashman, director of authorisation at the FCA, said the new regime will give consumers greater protections and firms a clear framework to operate in. The regulator said the framework reflects the government and FCA's determination that the UK's crypto sector should be defined by integrity and trust, along with growth and innovation.

The authorisation process is designed to establish that an applicant can meet FCA standards in practice, and policies that do not correspond to a firm's actual operating model are unlikely to be sufficient, according to JD Supra. Firms should compare their existing governance, financial resources, safeguarding, compliance, risk management, operational resilience and conduct arrangements against the FCA's final rules, and identify proposed senior managers with sufficient expertise and authority.

The FCA published its final crypto rules and guidance in June 2026, with the final rule package published on 30 June 2026 illustrating the scale of the transition, according to JD Supra. The regulator has confirmed the application of core handbook requirements to regulated cryptoasset firms and has published specific guidance on applying the consumer duty to the sector.

Conclusion

The opening of the FCA crypto authorisation gateway on 30 September 2026 starts the clock for crypto firms seeking to operate in the UK under full regulation. Existing businesses face a 28 February 2027 application deadline to benefit from transitional arrangements that let them continue operating while their applications are assessed, with the new regime taking full effect on 25 October 2027. Firms that cannot demonstrate they meet the FCA's standards on consumer protection, safeguarding, market integrity and financial resilience will not be authorised to operate in the UK market.

Frequently Asked Questions

When did the FCA crypto authorisation gateway open?

The FCA opened its authorisation gateway on 30 September 2026, allowing firms to apply for permission to carry on the new regulated cryptoasset activities created by the Financial Services and Markets Act 2000 (Cryptoassets) Regulations 2026. Firms can apply through FCA Connect for Part 4A permission.

What is the deadline for crypto firms to apply for FCA authorisation?

Firms intending to continue operating in the UK should apply by 28 February 2027, ahead of the new regime coming into force on 25 October 2027. The FCA has encouraged firms to apply as early as possible rather than treating the February date as a target.

Can existing UK crypto firms keep operating while their application is assessed?

Yes. Existing firms that apply during the application period can continue providing cryptoasset services, including taking on new business, while their application is assessed if a decision has not been made before the new regime comes into force on 25 October 2027, under statutory transitional provisions.

Does FCA registration under the Money Laundering Regulations carry over to the new regime?

No. A crypto business currently registered with the FCA under the Money Laundering Regulations will not automatically receive a Part 4A permission to conduct the corresponding activity from October 2027. Existing regulatory status does not automatically migrate into the new regime.

Do overseas crypto firms fall under the UK's new crypto regime?

Potentially. A firm whose principal place of business is outside the UK may still fall in scope where it has UK employees or agents undertaking regulated dealing or arranging functions, or operates through substantive UK-based operations, which could include UK servers, UK-targeted marketing or bank accounts.

Sources

Binance Delisting Knocks Epic Chain and QuickSwap Down as Much as 10% Within Minutes

Binance Delisting Knocks Epic Chain and QuickSwap Down as Much as 10% Within Minutes

by | Oct 14, 2026 | Binance

Binance said it will drop Epic Chain (EPIC) and QuickSwap (QUICK) from its platform, with every spot trading pair for the two tokens set to disappear on October 28, according to Yahoo Finance. The Binance delisting notice hit prices almost immediately, and QUICK plunged roughly 10% under the level it held on the exchange just before the announcement went live.

The decision carries weight for holders because tokens that lose a Binance listing usually see liquidity dry up, and that tends to push investors toward the exits before trading stops altogether. Both EPIC and QUICK had already spent months flagged under the exchange's Monitoring Tag, a label Binance reserves for assets it views as carrying more volatility and risk than the rest of its listings, and one it warns could eventually lead to a token failing its listing standards.

Tokens Whipsaw After Binance Delisting Notice

QUICK absorbed the heavier blow in the minutes that followed the announcement. Once the token had fallen roughly 10% from where it traded before the notice, the selling pressure persisted, and about 17 minutes later it was still changing hands 6.69% lower, according to TradingView data cited by Yahoo Finance.

EPIC traced out a very different path over the same stretch. The token gave up about 9% relative to its pre-announcement level, then clawed back the losses and more, trading 5.61% above that mark within the same window, Yahoo Finance reported. That snapback made EPIC an outlier, since Binance delisting notices have more often produced steep, sustained declines across recent rounds.

A Pattern of Steep Losses

Sharp selloffs in the wake of Binance delisting notices have become a recurring feature of the exchange's review cycles, according to Yahoo Finance. The August round delivered some of the clearest examples: ICON (ICX) tumbled 12.5% and Secret (SCRT) gave up 18.8% after the exchange posted its notice, declines that carried both tokens to fresh all-time lows.

A separate announcement earlier that same month produced similar damage, with PIVX losing 19.27% and Vulcan Forged PYR (PYR) dropping 18.31%. Not every token followed that script, though. Vanar (VANRY) managed to shake off its initial losses and finished the period up 8.23%, an outcome that closely mirrors the way EPIC bounced back on Wednesday.

Months Under the Monitoring Tag

Wednesday's Binance delisting did not arrive without warning signs. Both tokens had carried the exchange's Monitoring Tag for months before it moved to delist them, Yahoo Finance reported, with QUICK picking up the label on April 30 and EPIC added to the list on May 22.

Binance uses the tag to single out tokens it judges to be more volatile and risky than the rest of what it lists, and it cautions that any asset carrying the label could eventually fall short of its listing criteria. The exchange said the delistings followed its most recent reviews, a process in which it weighs factors that include trading volume, development activity, and team commitment.

Staged Wind-Down Begins This Week

Rather than pulling EPIC and QUICK all at once, Binance plans to phase the two tokens out of its product lineup in stages, beginning this week, according to Yahoo Finance. The first step lands on October 15, when the exchange cuts off margin borrowing for the pair and pulls them out of its Buy & Sell Crypto service.

The restrictions tighten on October 20. From that date, users will be blocked from opening new futures positions in either token, and Binance will settle every outstanding EPIC and QUICK contract. That same day, the exchange will close out existing loans tied to the tokens, strip them from margin trading, and remove them from Binance Pay.

On October 21, the cleanup reaches Spot Copy Trading, where both pairs will be delisted and any assets still sitting in those strategies will be force-sold at market price. Simple Earn follows the same day, redeeming Flexible and Locked positions in the two tokens back into users' Spot Accounts.

The spot market itself shuts down on October 28. At that point all open orders in the two pairs will be canceled and Trading Bots will stop running on them. Binance Convert and Gift Card will also drop the tokens, and starting October 29 the exchange will stop crediting EPIC and QUICK deposits to user accounts.

The final deadlines arrive at the end of the year. Binance will stop processing withdrawals of both tokens on December 28. After that, on December 29, the exchange may swap any leftover balances into stablecoins, though it said that conversion is not guaranteed, leaving holders who miss the withdrawal cutoff without certainty over what they will recover.

Conclusion

The Binance delisting of Epic Chain and QuickSwap caps months of both tokens sitting under the exchange's Monitoring Tag, and it extends a recent run of delisting notices that have triggered steep, immediate losses. QUICK dropped as much as 10% within minutes of the announcement, while EPIC recovered from an initial 9% slide to trade above its pre-announcement level. The next concrete step comes on October 15, when Binance cuts off margin borrowing for the two tokens and removes them from its Buy & Sell Crypto service, opening a staged wind-down that runs through late December.

Frequently Asked Questions

When will Binance delist Epic Chain (EPIC) and QuickSwap (QUICK)?

Binance will remove all spot trading pairs for EPIC and QUICK on October 28, canceling open orders and halting Trading Bots on both pairs. The wind-down starts earlier, with margin borrowing suspended on October 15, and withdrawals of the two tokens ending on December 28.

How much did QUICK and EPIC fall after the Binance delisting announcement?

QUICK fell about 10% below its pre-announcement level on Binance once the notice went live, and roughly 17 minutes later it was still trading 6.69% lower, according to TradingView data. EPIC slid about 9% before rebounding within the same window to trade 5.61% above its pre-announcement mark.

Why is Binance delisting EPIC and QUICK?

The exchange pointed to its most recent reviews as the basis for the move, with trading volume, development activity, and team commitment among the factors it weighs. Both tokens had also spent months carrying Binance's Monitoring Tag, a label the exchange applies to assets it considers more volatile and risky than its other listings.

What happens to EPIC and QUICK balances left on Binance?

Holders have until December 28 to move their tokens off the exchange, which is when withdrawal processing for both assets ends. Starting December 29, leftover balances may be swapped into stablecoins, but Binance cautioned that this conversion is not guaranteed, so anyone who misses the cutoff has no certainty over what they will get back.

Have other tokens crashed after Binance delisting notices?

Yes. In August, ICON (ICX) fell 12.5% and Secret (SCRT) lost 18.8% after the exchange's notice, with both sinking to fresh all-time lows. Earlier that month, PIVX dropped 19.27% and Vulcan Forged PYR shed 18.31%, while Vanar (VANRY) was an exception, finishing 8.23% higher despite early losses.

Sources

ASML AI Demand Forecast Sees Chip Equipment Shortfall Through 2028

ASML AI Demand Forecast Sees Chip Equipment Shortfall Through 2028

by | Oct 14, 2026 | How to Crypto

ASML, the Dutch maker of semiconductor manufacturing equipment, said on Wednesday that demand for its chipmaking systems driven by artificial intelligence investment will outstrip its supply in 2027 and 2028, even after planned production increases. The company said customers are increasingly securing equipment well beyond the next two years as chipmakers race to build capacity for AI-related applications.

The warning matters because ASML's lithography systems occupy a critical position in the global semiconductor supply chain, where they transfer detailed circuit designs onto silicon wafers as chips are made. The company's extreme ultraviolet machines are especially vital for manufacturing some of the most advanced processors available, and the tools are so complex and costly to produce that supply cannot be raised quickly even when orders surge.

ASML AI Demand Extends to 2030

Chief Executive Officer Christophe Fouquet said customer discussions have moved well past the company's near-term planning horizon, with order talks now routinely bypassing near-term delivery slots.

"A lot of our discussions with our customers now are shifting in fact beyond 2027," Fouquet said in a statement. He said the company was receiving substantial orders for 2028 delivery, alongside growing interest from customers seeking to secure equipment for 2029 and, in some cases, 2030.

According to Yahoo Finance, Fouquet said clients kept lifting their equipment requirements across both near-term and medium-term horizons, a shift he attributed mainly to expectations about how large and how long AI-related spending will run. The comments suggest chipmakers are making longer-term commitments to lock in access to the specialised systems needed to expand production.

Capacity Plans May Fall Short

ASML announced in July that it would expand output by 30 per cent in 2027 and then add another 30 per cent the following year. The latest demand outlook indicates that even those expansions may not be sufficient to fulfil all customer requirements, and the company is examining additional ways to raise output.

Whether ASML can grow beyond those plans hinges on what its factories can physically build, how much investment is required and how long scaling up takes. Executives are weighing further production options to ease assembly constraints ahead of the deadline for 2028 delivery pledges, Retail News Asia reported.

Retail News Asia reported that intense appetite for advanced lithography machines has pushed delivery timelines far beyond customary waiting periods, dragging commercial negotiations into the next decade. Chipmakers in Taiwan, South Korea and Japan cannot accelerate output at new wafer fabrication plants faster than the machinery arriving allows, and tool allocations dictate foundry expansion schedules.

Fourth-Quarter Forecast Tops Estimates

ASML guided fourth-quarter 2026 revenue to a range of 14.2 billion euros to 15.2 billion euros, equivalent to $15.9 billion to $17.0 billion, and projected a gross margin near 57 per cent. That top-line outlook came in above the 14.19 billion euro consensus forecast from analysts surveyed by LSEG.

Third-quarter net sales came in at 11.66 billion euros, up from 9.33 billion euros in the prior quarter, according to Yahoo Finance, keeping ASML the largest chipmaking equipment supplier globally. Profit climbed to 4.96 billion euros, compared with 2.92 billion euros in the preceding quarter, and gross margin improved to 57.5 per cent from 54 per cent.

ASML also raised its full-year 2026 sales forecast on Wednesday after the record third-quarter results, as soaring AI investment boosted demand for semiconductor manufacturing equipment, Yahoo Finance reported. The company lifted its annual net sales target to a range of 44 billion euros to 45 billion euros, narrowing the prior guidance of 43 billion euros to 45 billion euros.

Pricing Power and Market Read-Through

Should ASML fall short of AI-driven orders in 2027 or 2028, buyers would be left vying for scarce production slots, a dynamic that usually shifts leverage to the supplier, according to Finimize. Scarcity makes it easier to keep prices firm, limit discounts and prioritise higher-value systems and services, all of which supports margins, the outlet said.

Finimize said the 57 per cent margin target offers a window into who holds pricing power, and that attention may turn away from quarterly order fluctuations toward whether ASML can ramp output on schedule and convert multiyear commitments into delivered machines and steady profits.

Crypto Briefing reported that market pricing implies traders treat the ASML outlook as a tailwind for the valuations of AI firms such as Anthropic. The sustained demand signal appears to boost confidence in AI-related sectors, the outlet reported.

What the Industry Watches Next

Market participants will closely monitor ASML's capacity expansion plans and their potential impact on the supply chain, according to Crypto Briefing. How ASML divides up its machines among buyers, and whether it revises the planned 30 per cent annual capacity increases, will set the pace at which semiconductor plants worldwide receive their tools, Retail News Asia reported.

Semiconductor manufacturers seeking to expand output may have to plan further ahead to secure equipment, while ASML must balance capacity expansion against the technical complexity of its manufacturing operations. Because buyers are already asking about deliveries stretching to the close of the decade, ASML must decide how much extra output it can add without compromising production standards.

ASML's next phase of expansion will be closely watched by the semiconductor industry as manufacturers seek to meet the computing demands of the growing AI market. The widening gap between customer demand and available production capacity presents a challenge for both ASML and its customers.

Conclusion

ASML has established that AI-driven demand for its chipmaking systems will exceed its supply in 2027 and 2028, even with planned 30 per cent annual capacity increases, and that customers are already negotiating for 2029 and 2030 delivery slots. The company backed the outlook with a fourth-quarter 2026 sales forecast of 14.2 billion euros to 15.2 billion euros, above analyst estimates, and a raised full-year sales target of 44 billion euros to 45 billion euros.

The next concrete step is ASML's review of additional manufacturing options to expand 2028 capacity beyond the previously communicated 30 per cent increase. The industry will watch for decisions on equipment allocations and any revision to those capacity plans, which will set delivery timelines for semiconductor plants worldwide.

Frequently Asked Questions

What did ASML forecast for AI chip equipment demand in 2027 and 2028?

ASML said on Wednesday that demand for its chipmaking systems, driven by artificial intelligence investment, will exceed its supply in 2027 and 2028. The shortfall persists even after planned production capacity increases of 30 per cent in each of those years, and the company is examining additional ways to raise output.

What is ASML's fourth-quarter 2026 sales forecast?

ASML forecast fourth-quarter 2026 sales of between 14.2 billion euros and 15.2 billion euros ($15.9 billion to $17.0 billion), with a gross margin of approximately 57 per cent. The revenue forecast exceeded the 14.19 billion euro average estimate of analysts polled by LSEG.

How far ahead are ASML customers placing orders?

CEO Christophe Fouquet said ASML is receiving substantial orders for 2028 delivery, alongside growing interest from customers seeking to secure equipment for 2029 and, in some cases, 2030. He said customer discussions are shifting beyond 2027 as chipmakers lock in access to specialised lithography systems.

How much does ASML plan to expand production capacity?

In July, ASML announced plans to increase production capacity by 30 per cent in 2027 and by a further 30 per cent in 2028. The latest demand outlook indicates those expansions may not fulfil all customer requirements, so the company is investigating ways to expand 2028 capacity beyond the previously communicated plan.

Why does the ASML supply shortfall matter for chipmakers?

ASML's lithography systems are critical to chip manufacturing, and its extreme ultraviolet technology is used for the world's most advanced processors. For chipmakers in Taiwan, South Korea and Japan, machinery shortfalls limit how fast new wafer fabrication plants can ramp output, and tool allocations dictate foundry expansion schedules.

Sources

Google DeepMind, GovTech Unveil On-Device Scam Detection Tool for Singapore

Google DeepMind, GovTech Unveil On-Device Scam Detection Tool for Singapore

by | Oct 14, 2026 | Fight the FUD

Google DeepMind and Singapore's Government Technology Agency (GovTech) on October 14, 2026, announced an on-device scam detection tool built to catch conversational scams inside private messaging apps, issuing real-time warnings without sending user data to the cloud. Minister for Digital Development and Information Josephine Teo unveiled the strategic collaboration at Singapore International Cyber Week (SICW).

The system is designed to catch manipulation as a conversation unfolds, warning users before they transfer money, disclose credentials or share other sensitive information. Because the analysis runs entirely on the handset, the partners say it can work even inside end-to-end encrypted messaging environments where platform operators cannot inspect private chats on a central server.

How the On-Device Scam Detection Model Works

The tool uses a lightweight version of Google's Gemma model that was fine-tuned on Singapore-specific scam patterns, per a Google DeepMind blog post written by JJ Geewax and Miguel de Andrés. Training material covered impersonations of public institutions, banks and local figures, as well as trending regional fraud schemes, so the model can distinguish harmless social chatter from genuine manipulation.

To keep the classifier practical on consumer smartphones, the teams built a two-stage pipeline. A low-power classifier first screens inbound messages using minimal energy, and conversations that look suspicious are escalated to the fine-tuned Gemma model, which evaluates a larger snapshot of the exchange to confirm a threat or discard a false alarm. Battery efficiency was a central design concern, the announcement said.

If the system spots signs of manipulation in a chat, it triggers an on-screen alert telling the user to hold off before sending money or sensitive details. The goal is early intervention, ideally before funds move, and the resulting model is optimized to run on typical smartphones used in Singapore.

Privacy Design and Governance

The partners framed privacy as the project's defining constraint. Conventional threat detection scans messages on cloud servers, but many messaging platforms are end-to-end encrypted and cannot inspect private conversations, the announcement explained. Running the analysis locally keeps conversation data on the device, with no transmission to an external cloud service.

"Fighting scams shouldn't mean sacrificing personal privacy," said Mark Chen, Director of Trust and Safety at GovTech Singapore.

The announcement describes governance measures it labels Sovereign Sandbox, No Citizen Data and Stress-Testing. Fine-tuning took place inside a GovTech-managed and maintained Google Cloud Platform environment that kept training data isolated, private or personally identifiable information was strictly excluded in favor of anonymised datasets drawn from scam reports and intelligence feeds, and the model was validated using synthetic conversations and local language samples spanning multiple fraud types and shifting impersonation methods.

Why Singapore Is Targeting Conversational Scams

According to the announcement, more than half of all reported crimes in Singapore are scams, with losses exceeding S$4 billion since 2019. The partners say tactics have evolved beyond basic phishing into sophisticated social engineering, where attackers construct believable dialogues to manipulate victims before requesting payment, and such schemes now account for up to 80% of all losses.

Teo positioned the collaboration within a wider effort to combat AI-powered scams, observing that the challenge has intensified as fraudsters adopt increasingly advanced AI tools. Generic models often struggle with localised deception techniques, the announcement said, which is why the Gemma model was tuned on Singapore-specific material.

The tool adds to an existing stack of government defenses. SATIS, the Scam Analytics and Tactical Intervention System, operates for the Singapore Police Force and takes down roughly 20,000 malicious websites each month, while GovTech's rMSE system scans over 600,000 websites daily and ScamShield provides another layer of protection. The new model extends that apparatus into multi-turn conversational detection.

Rollout Questions and Platform Integration

Teo did not give a launch date for the Gemma-based tool and did not disclose performance metrics like detection or false-positive rates. The model is optimised for smartphones commonly used in Singapore, but the partners have not said whether it will be built into Android or particular messaging services, according to CXO Digitalpulse.

The announcement says messaging platforms and operating system vendors can directly integrate the lightweight models to safeguard users locally, and it presents the project as a template for balancing user protection with end-to-end encryption. A detection model only helps if it is built into the apps people actually use, which typically requires cooperation from messaging platforms and device makers.

The work was delivered under Google's National AI Partnership with the Singapore Government, which Google DeepMind announced on May 20, 2026, introducing new country programmes under its National Partnerships for AI initiative following the opening of a Singapore research lab the previous year. The collaboration was presented at Singapore International Cyber Week 2026, where the opening address was delivered on October 13 by Senior Minister and Minister for Home Affairs K Shanmugam, according to the address published by the Ministry of Home Affairs.

Accuracy and Adaptation Challenges

The project faces open questions beyond timing. A detector that raises false alarms too frequently risks being tuned out by users, while one that overlooks real threats provides little meaningful defense, so accuracy will determine whether warnings change user behavior.

Fraudsters also shift their methods rapidly, and Teo herself noted that AI is complicating efforts to stop them. A model trained on current scam scripts will require ongoing retraining to remain effective against new ones, a maintenance burden the partners have not yet detailed publicly.

Conclusion

Google DeepMind and GovTech have built a lightweight, Gemma-based model that flags manipulative scam conversations on the device itself, fine-tuned on Singapore-specific schemes and governed by rules that keep citizen data out of training. Announced at Singapore International Cyber Week on October 14, 2026, the on-device scam detection system is optimized for common local smartphones and designed for end-to-end encrypted messaging. The next concrete step is integration: the partners say messaging platforms and operating system vendors can embed the models directly, though no launch date or performance figures have been disclosed.

Frequently Asked Questions

What did Google DeepMind and GovTech Singapore announce on October 14, 2026?

They announced an on-device AI model, built on a lightweight version of Google's Gemma, that detects manipulative scam conversations in private messaging apps and warns users in real time. Minister Josephine Teo unveiled the collaboration at Singapore International Cyber Week, with analysis running entirely on the user's device.

How does the Gemma-based scam detection model protect user privacy?

The model processes conversation data on the handset rather than sending it to a cloud server, so it can operate inside end-to-end encrypted messaging apps. Training excluded private or personally identifiable citizen data, using anonymised material from scam reports and intelligence feeds inside a GovTech-managed Google Cloud environment.

Why is Singapore investing in conversational scam detection?

Scams make up over half of Singapore's reported crimes and have cost more than S$4 billion since 2019, according to the announcement. Social-engineering schemes, where attackers build credible conversations before requesting money, now account for up to 80% of scam losses, and generic models often miss localised deception techniques.

When will the on-device scam detection tool launch in Singapore?

No public release date has been disclosed. Minister Josephine Teo did not provide a launch timeline or performance figures such as detection and false-positive rates, and details of integration into Android or specific messaging platforms remain undisclosed, according to CXO Digitalpulse.

How does the two-stage pipeline save smartphone battery life?

A lightweight, low-power classifier first screens incoming messages for potential threats using minimal energy. Only conversations flagged as suspicious move to the fine-tuned Gemma model, which analyses a broader message sequence to confirm manipulation or dismiss a false alarm, balancing detection capability with smartphone processing and battery constraints.

Sources

XBOW Chrome Full-Chain Exploit Built With GPT-5.6 Earns $250,000 Google Bounty

XBOW Chrome Full-Chain Exploit Built With GPT-5.6 Earns $250,000 Google Bounty

by | Oct 14, 2026 | How to Crypto

Autonomous offensive security company XBOW said on October 13, 2026 that Google paid it a $250,000 bounty after its researchers, assisted by AI agents powered by OpenAI's GPT-5.6 Sol model, produced a full-chain exploit of Chrome. The Seattle-based startup said the payout marks the first $250,000 Chrome full-chain reward granted since Google's updated Vulnerability Reward Program rules took effect in April 2026, and CEO Oege de Moor wrote on X that it was only the second time in Chrome's history Google had paid that amount for a full chain.

The XBOW Chrome full-chain exploit combined three previously unknown vulnerabilities in Chrome for Windows and could have allowed an attacker to execute arbitrary code beyond the browser's sandbox protections, set off by a victim simply loading a web page. XBOW estimates the flaws affected roughly one billion Chrome users on Windows globally, a figure FourWeekMBA noted is the company's own estimate.

Six days in Las Vegas

The team built the exploit between August 2 and August 7, 2026, while in Las Vegas for the Black Hat and DEF CON security conferences, according to a blog post XBOW published on October 13. Researcher Brendan Dolan-Gavitt worked with the AI agents through the campaign, with de Moor directing the effort.

XBOW put the compute bill for the six-day campaign at roughly $40,000 in API charges, a reconstruction Dolan-Gavitt produced from archived Codex logs. That is a fraction of the $250,000 payout, and the company said that in the era before large language models, a four-person team would typically need somewhere between six and nine months to assemble a comparable Chrome full chain. XBOW's first submission attempt took two months, Dolan-Gavitt wrote, and Google told the team it already knew about one of the bugs in that first chain.

De Moor wrote that a veteran of a three-letter intelligence agency told him that identifying this class of vulnerability and gaining total access once took his team half a year. The model involved, GPT-5.6 Sol, is publicly available rather than a restricted lab tool, a point de Moor said was deliberate: the team picked a model anyone can access, not a specialized security model.

How the three-bug chain worked

Chrome is built in layers, so an attacker who finds a single bug might take over a web page but get no further. A full chain strings multiple bugs together so that each one unlocks the next barrier, carrying the attacker from an ordinary web page all the way to the underlying computer. XBOW's chain crossed two sandbox boundaries in three stages on Windows.

The first stage exploited a bug in the Maglev compiler inside V8, Chrome's JavaScript engine, where a mutable boxed number was reused during deoptimization, handing the exploit read and write reach across V8's pointer-compression cage, the first layer of Chrome's security architecture. The second bug, tracked as CVE-2026-85045, sat in Chrome's animated WebP image decoder, where cached geometry data did not match changed image bytes, producing one precisely targeted write to native memory outside the cage.

The final stage hinged on CVE-2026-78952, a flaw in Crashpad, the component Chrome uses to handle crashes, which treated a failed read from a supposedly frozen renderer process as a stable, trustworthy observation. Abusing it converted the second stage's native memory write into unrestricted code execution beyond the sandbox. In the end, code ran outside Chrome's sandbox with no downloads, no clicks and no user interaction beyond navigating to a page.

Viewed individually, each of the three flaws might have been rated only low or moderate in severity, according to SMBtech. Linked in sequence, they yielded a full escape from the sandbox, a class of browser vulnerability that security researchers and attackers alike rank among the most severe.

Humans set the strategy

XBOW's own account describes a division of labor between the researchers and the agents. The agents combed through Chrome's source code, ran experiments on candidate ideas, assembled proofs of concept and drafted the first versions of the exploit, while the researchers chose targets, supplied context, checked the agents' work and redirected them.

According to the company, a single judgment call by the humans made the difference: the exploit was assembled from three separate modules, and each module had a backup option, so any bug that turned out to be already known could be replaced without starting over. Dolan-Gavitt wrote that this design rescued the chain when its first V8 sandbox escape turned out to be a duplicate. Alongside Dolan-Gavitt, researchers Andres Luksenberg, Baul Lee and Willem Pinckaers took on the challenge.

XBOW said it believes that within a year each of the three bugs could be found by agents working on their own. Dolan-Gavitt cautioned that the outcome does not put a Chrome chain within reach of just anyone willing to spend that much time and money, though in his view the barrier has dropped considerably. He also wrote that bug collisions are a major risk now that everyone, Google included, is pointing LLMs at codebases like Chrome.

Google's response and the disclosure

XBOW reported all three bugs to Google through the official vulnerability reward program, and Google confirmed the findings and paid the reward about a month after submission. Google went on to reproduce the exploit itself, verify the vulnerabilities and release patches, and none of the flaws were ever turned against Chrome users. Public root-cause reports for all three vulnerabilities have been made available.

De Moor framed the disclosure in terms of what the combination of human expertise and AI capability now makes possible in offensive security research, according to SMBtech. He pointed to the timeline and cost as indicative of a broader shift in the economics of vulnerability research, and argued that restricting access to powerful models is the wrong question because the model XBOW used is already public.

OpenAI president and co-founder Greg Brockman quoted de Moor's post on X the same day, restating the six days, $40,000 of compute with GPT-5.6 Sol and the estimated one billion affected users, and writing that it highlights the opportunity for defenders to use these tools to secure their software. Clint Gibler, OpenAI's head of cyber research, highlighted the defensive implications of the work, SMBtech reported.

What XBOW does next

XBOW's stated position is that defenders need to match the speed at which attackers can now operate, and that autonomous offensive security platforms are a way to do that. The company said the tools that found these bugs are available to anyone who defends software, and that defenders have an advantage attackers never will: they can test their own code first.

XBOW said it will use the $250,000 award to fund efforts to secure non-governmental organizations using the company's technology. It is also developing an autonomous product for testing compiled software, such as desktop applications, servers and embedded devices, and partnering with large organizations to design that product.

XBOW was founded in 2024 by de Moor, who previously created GitHub Copilot. According to the company, it has submitted more than 1,060 vulnerabilities across real-world production targets. XBOW does not give a total cost for the Chrome campaign including researcher time, and it has not said how much compute went into the first, rejected chain.

Conclusion

XBOW's disclosure establishes that a small team pairing human researchers with publicly available AI agents can produce a Chrome full-chain exploit in days rather than months, and that Google will pay its top tier for the result. The three vulnerabilities are patched, the $250,000 bounty is paid, and the root-cause reports are public. The next concrete steps are XBOW's planned use of the award to fund security work for non-governmental organizations and its development of an autonomous product for testing compiled software such as desktop applications, servers and embedded devices.

Frequently Asked Questions

How much did Google pay XBOW for the Chrome full-chain exploit?

Google paid XBOW $250,000 for the full-chain exploit. XBOW says it is the first award of its kind under Chrome Vulnerability Reward Program rules that took effect in April 2026, and CEO Oege de Moor says it is only the second time Google has paid that amount for a Chrome full chain.

What were the three Chrome vulnerabilities XBOW chained together?

The chain combined a bug in the Maglev compiler inside V8, Chrome's JavaScript engine; a flaw in the animated WebP image decoder tracked as CVE-2026-85045; and a memory safety issue in Crashpad, Chrome's crash handler, tracked as CVE-2026-78952. Together they produced a full sandbox escape on Windows.

How long did the XBOW Chrome exploit take and what did it cost?

The team built the exploit in six days, from August 2 to August 7, 2026, during the Black Hat and DEF CON conferences in Las Vegas. XBOW reported compute costs of around $40,000 in API fees, reconstructed from archived Codex logs, a fraction of the $250,000 payout.

Were Chrome users attacked with these vulnerabilities?

No. XBOW reported all three bugs to Google through the official vulnerability reward program, the flaws were never used against Chrome users, and Google reproduced the exploit, confirmed the vulnerabilities and shipped fixes. Public root-cause reports for all three vulnerabilities have been made available.

Which AI model did XBOW use to find the Chrome bugs?

XBOW's agents ran on OpenAI's GPT-5.6 Sol model, which is publicly available rather than a restricted or specialized security model. De Moor said the team deliberately chose a model anyone can access, and he argues that restricting powerful models is the wrong question because the model is already public.

Sources