Singapore AI news archive.
Every SG + AI news item we classified, grouped by ISO week. Each card carries the analyst one-liner, category tag, and sentiment dot that appeared on the live reel. Kept permanently so we can look back on how the ecosystem evolved.
About this archive: each item is ingested and classified automatically every day with a category (policy, enterprise, funding, talent, infrastructure or products), a sentiment read, and a one-line analyst note. The archive is kept permanently. The same feed powers the news sections on the altronis.sg homepage and the @AIinSG Telegram channel.
Week 42 · 2026
60 items
PolicyTNG · Ping An Digital Bank reports over 96% deepfake detection in HKMA sandbox
HKMA's sandbox is forcing banks to prove AI can catch AI fraud, setting a compliance bar that will ripple across the region.
PolicyVERGE · Trump posts AI-edited photo to brag about White House restoration work
SynthID flagged a White House post, proving watermarking works but also that high-level users ignore provenance.
PolicyWIR · AI Hacks Leave a Mess That AI Companies Aren’t Paying For
Regulatory deference to innovation is letting AI firms externalize the cost of agent misbehavior onto victims, a gap that will persist until enforcement catches up.
PolicyCNBC · Anthropic taps nonprofits METR and SecureBio to help assess risks and safety practices
Anthropic is operationalizing third-party audits with a tiered access model, setting a precedent for how frontier labs will structure external safety oversight.
PolicyTC-AI · Anthropic is letting METR researchers look under the hood with employee-level access
Independent oversight is becoming a structural requirement, not a PR gesture, as labs embed external evaluators to manage regulatory risk.
PolicyCNBC TV · Watch CNBC's full interview with Nebraska Attorney General Mike Hilgers
Eight-state coalition targets consumer safety claims, forcing OpenAI to defend its minor-protection architecture in court.
PolicyARS · Report: FTC probes Anthropic, OpenAI after Trump dismissed AI fears as “hoax”
Regulators are aligning AI incident reporting fines with data breach standards, forcing labs to treat security disclosures as a compliance liability.
PolicyCNBC TV · West Virginia AG says OpenAI knew of flaws that put children at risk
State-level litigation is now targeting specific safety failures, forcing labs to document risk mitigation for minors.
PolicyVERGE · Nobody is safe from celebrity AI deepfakes
Verified badges no longer guarantee authenticity, forcing platforms to retrofit identity verification before deepfakes become the default news cycle.
PolicyCNA · Small states must have say as world shapes new compact with tech giants: Vivian Balakrishnan
SG is positioning itself as the bridge between small states and tech giants, pushing for a global compact that prevents corporate monopoly over AI governance.
PolicyCNBC · Nvidia must work with Chinese AI companies on safety, Jensen Huang tells CNBC
Huang pushes for bilateral safety channels, positioning Nvidia as the bridge between US and Chinese AI governance.
PolicyCNBC TV · 'One puts AI as a threat; the other puts AI as a target': Singapore minister on cyber defences
SG leadership frames AI as both threat and target, signaling a dual-track cyber defense strategy for the sector.
PolicyST · Rules governing tech can’t be monopolised by a handful of powerful states, big tech firms: Vivian
SG is positioning itself as a bridge between frontier labs and small states, betting that consensus norms will outlast unilateral control.
PolicySCMP · China, US discuss AI as chip stock rout hits Asian markets
A formal incident hotline between the two largest AI powers suggests the geopolitical risk premium for compute supply is finally being priced in.
PolicyST · OpenAI backs Singapore’s call for AI incident transparency
OpenAI’s vague commitment to 'immediate' alerts without a fixed timeline leaves SG regulators with a compliance gap they must now close.
PolicyCNA · AI accelerating cyber threats, making regional cooperation crucial: Australian minister
AI compresses attack timelines from weeks to seconds, forcing SG-AU to tighten critical infrastructure defenses.
PolicySCMP · AI job losses set to rein in office demand in developed Asia-Pacific economies: report
CBRE data suggests AI-driven job displacement will trim 5% from office demand growth, signaling a structural shift in SG real estate.
PolicyBT · Nvidia CEO says he would cut off buyers who build unsafe AI, bubble fears overdone
Nvidia’s $250m UK research pledge and supply-cut threat position the chip giant as the de facto safety gatekeeper for the AI stack.
PolicyST · Nvidia CEO says he would cut off buyers who build unsafe AI
Nvidia's $250m UK safety pledge and threat to cut off unsafe buyers signals a shift toward vendor-enforced compliance.
PolicyVERGE · Eight states sue OpenAI, saying ChatGPT isn’t safe for kids
Coordinated state suits force OpenAI to prove safety claims, raising the compliance bar for all consumer-facing AI.
PolicyCNBC · OpenAI sued by 8 states over ChatGPT safety concerns
State-level liability is now the primary regulatory vector, forcing OpenAI to retrofit compliance for a fragmented US market.
PolicyCNBC TV · Iowa, Kansas and 6 other states reportedly sue OpenAI over claims that ChatGPT harms children
State-level litigation against OpenAI signals a shift from federal oversight to fragmented legal pressure on consumer AI safety.
PolicyCNBC TV · OpenAI escalates trade secret fight with Apple
Legal friction between top labs is now a structural risk for hardware partnerships, not just a PR blip.
PolicyVERGE · Now that Character.AI is going bankrupt, what happens to all of its users’ data?
Forty state AGs are forcing data protection terms into the bankruptcy sale, setting a precedent for how AI chatbot data is handled in insolvency.
PolicyARS · Recording device hidden in toilet paper found in top-secret OpenAI meeting room
Security breaches at top labs now trigger federal espionage charges, raising the compliance bar for AI infrastructure.
PolicyCNBC TV · Anthropic urges court to rehear appeal of Pentagon blacklisting
Anthropic is escalating its legal challenge against the Pentagon, signaling that AI procurement disputes are now high-stakes court battles.
PolicyTNG · US-based Anthropic opens Singapore office, signs MOU with the Singapore government
The MOU binds a top lab to local safety standards, signaling that regulatory rigor is now a core requirement for enterprise AI deployment.
PolicySCMP · AI risks could repeat China shock that cost jobs, former US commerce chief Raimondo warns
Raimondo frames AI as a structural labour shock, pushing policymakers to plan transitions rather than react to disruption.
PolicyARS · Anthropic warns small AI models could self-replicate through IoT devices
Anthropic flags a new worm vector where small models self-replicate via IoT, forcing a rethink of edge security.
PolicyDEC · Alleged OpenAI spy vibe-coded a surveillance device using his wife's chatbot account
FBI flags AI IP theft as a top counterintelligence priority, raising the compliance bar for firms with Gulf-state capital ties.
PolicyBT · Bubble or nothing? Looking at the AI boom as part of a cycle
MAS frames AI as a structural cycle rather than a bubble, forcing SG planners to build resilience into economic strategy.
PolicySCMP · China approves Alibaba AI models for use in iPhone Duo and other Apple handsets
China’s regulatory gatekeeping forces Apple to localize AI with Alibaba, reshaping the global model distribution landscape.
PolicyARS · OpenAI, Google pushing new AI models easily tricked into making nudes, researchers say
Major labs are shipping models with exploitable safety gaps, forcing regulators to tighten oversight before widespread misuse.
PolicyCNBC TV · Justice Department says Saudi national arrested for installing audio recording device at OpenAI
Physical security breaches at top labs are now a federal matter, raising the compliance bar for all frontier AI facilities.
PolicyCNA · Anthropic opens Singapore office, signs AI safety agreement with government
Frontier labs are now subject to local safety evaluation, shifting the compliance burden from voluntary to structural.
Policye27 · Anthropic’s Singapore MOU bets on safety testing, not consumer reach
SG is positioning as a safety testing hub, not a consumer market, which shifts the value proposition for local AI firms.
PolicyTNG · Vietnam should review foreign-ownership caps, citing Western firms’ past regrets over JV
US pressure to lift foreign caps is a direct bid to secure Vietnam as the region's compute hub, linking trade access to AI infrastructure.
PolicyBT · Anthropic works to bring frontier AI models ‘on soil’ in Singapore, signs government pact
A government MOU on safety and evaluation signals that frontier model deployment is now a sovereign compliance issue, not just a product launch.
PolicyWIR · AI Models Can Be Easily Modified to Assist Terrorist and Extremist Activity
The speed of abliteration exposes a critical gap in current safety frameworks, forcing a pivot from existential risk to immediate misuse.
PolicyST · Anthropic signs MOU with Singapore to evaluate AI models for safety and security
MDDI is institutionalising model evaluation, positioning SG as the regional hub for frontier AI safety compliance.
PolicyST · Anthropic to work with Singapore to evaluate AI models for safety and security
MDDI and Anthropic formalize model evaluation, positioning SG as the regional hub for frontier AI safety compliance.
Policye27 · Singapore’s next AI industry may be the people who check the agents
IMDA is formalizing the verification layer, turning agent compliance from a theoretical risk into a billable service line.
PolicyCNA · Singapore’s AI-led economic growth set to continue in 2027: MAS
MAS confirms AI is now the primary engine of SG growth, with tech sectors driving 75% of expansion despite moderating pace.
PolicyCNA · Singapore forms task force to strengthen defences against AI-enabled cyber threats
IMDA is shifting from app-level testing to systemic resilience, forcing SG firms to rethink network segmentation for agentic threats.
Policye27 · Singapore wants Google’s AI to catch scammers before victims hit send
SG is embedding AI directly into the consumer interface to intercept fraud, shifting the burden from post-hoc recovery to real-time prevention.
PolicyST · Singapore launches new tool kit to help firms test AI agents for safety and reliability
IMDA is formalising agent evaluation standards, forcing SG firms to build audit trails before deploying autonomous workflows.
PolicyCNA · GovTech and Google DeepMind developing real-time detection of AI scams on devices
On-device AI detection shifts the scam defense burden from centralized servers to user hardware, prioritizing privacy.
PolicyWIR · An OpenAI Employee Allegedly Used a Chatbot to Help Plant a Recording Device at the Office
AI tools are now being weaponized for physical intrusion, forcing labs to rethink internal security beyond digital access controls.
PolicyCNBC · Anthropic urges court in Washington, D.C. to rehear appeal of Pentagon blacklisting
The Pentagon's blacklist of Anthropic models is now a legal test of whether safety refusals can trigger national-security sanctions.
PolicyCNBC · FBI arrests Saudi OpenAI employee who allegedly planted recording device in conference room
Physical security breaches at top labs are now a regulatory flashpoint, forcing a rethink of physical access controls.
PolicyTC-AI · Judge orders OpenAI to turn over hiring records in Apple trade secrets case
Legal friction is now a direct cost of the AI hardware race, forcing OpenAI to expose its talent acquisition playbook to Apple.
PolicyWIR · OpenAI and Anthropic Tell British Lawmakers: Regulate Us
Frontier labs are publicly splitting on statutory mandates, forcing regulators to pick sides rather than rely on voluntary pledges.
PolicyMS · ‘What matters is whether SMEs can put AI to good use’: Josephine Teo
Removing cost barriers for 10k SMEs shifts the bottleneck from access to actual workflow integration and upskilling.
PolicyCNBC · Michael Dell says AI needs safeguards and controls, and government has a role to play
Dell's call for bounded permissions and government oversight signals that enterprise leaders are now actively shaping the regulatory narrative.
PolicyDEC · 406 scientists demand AI models be cut off from the internet immediately
The push for mandatory airgapping and international oversight signals that containment failures are now a regulatory priority, not just a technical bug.
PolicyMS · LTA pauses pilot for self-driving public buses, will consider use for more targeted commuter groups first
LTA is narrowing the scope of autonomous bus pilots to employee shuttles, a pragmatic step that lowers regulatory risk while keeping the tech in the ecosystem.
PolicyVERGE · Actress Mia Goth says it’s ‘incomprehensible’ that Instagram allowed an AI image of her to spread
Meta's refusal to ban AI likenesses of real people is creating a liability gap that platforms are now forced to address.
PolicyCNBC TV · IMF Deputy Director warns AI debt boom could trigger a global financial correction
IMD officials are flagging AI leverage as a systemic risk, suggesting the debt-fueled buildout may face a hard correction.
PolicyBT · IMDA expands digital support for SMEs
IMDA is shifting from generic digital grants to curated AI tooling, aiming to AI-enable 70% of pre-approved solutions within two years.
PolicyST · More support for SMEs to adopt AI with new industry tie-ups, curated list of over 100 tools
IMDA is shifting from generic grants to curated tooling and data prep, addressing the messy data reality that blocks SME AI adoption.