12 May 2026: GM Reshapes for AI, Helsing Raises $1.2bn, and OpenAI Takes the Consulting Route (AM)

This morning’s AI news is defined by the business of artificial intelligence as much as the technology itself. A major car manufacturer is restructuring its IT workforce around AI skills, OpenAI’s consulting arm is building enterprise moats on the Palantir model, Europe’s defence AI sector just secured its largest round yet, and a new cybersecurity startup is sounding the alarm about AI-powered social engineering attacks that any UK business can face today.

General Motors has laid off hundreds of IT workers and is actively hiring replacements with stronger AI skills, one of the clearest signals yet that large corporations are moving from AI experiments to full-scale workforce restructuring. The new GM roles focus on AI-native development, data engineering and analytics, cloud engineering, and agent and model development. Prompt engineering and AI workflow design also feature on the hiring list, reflecting how much the day-to-day work of building software has shifted in the past year.

For UK workers and businesses, the message is pointed. The most protected roles appear to be those with hands-on AI experience, whether building workflows, evaluating model outputs, or managing AI tooling in production. Roles involving repetitive IT processes or legacy application maintenance are under the most pressure. If a company the size of GM is restructuring hundreds of positions at once, smaller organisations face compounding pressure to upskill or rehire around the same capabilities.

OpenAI’s internal deployment subsidiary, known as DeployCo, is taking a deliberately Palantir-style approach to enterprise AI, building bespoke workflows for large clients that are designed to be difficult for rival AI providers to replicate. Reporting from The Decoder today describes the Deployment Company as a majority-controlled OpenAI subsidiary, set up to embed deeply within client organisations and turn AI integration into durable, workflow-specific implementations rather than off-the-shelf subscriptions.

Palantir built much of its enterprise business by weaving software so tightly into client operations that switching became prohibitive. OpenAI appears to be applying the same logic to commercial AI deployment. For UK businesses currently evaluating AI deployment partners, this changes the landscape in two ways: OpenAI is now competing directly with the consultancies and systems integrators that have traditionally handled this work, and any DeployCo engagement comes with the same lock-in questions that apply to any deeply embedded vendor relationship.

German AI defence startup Helsing is set to raise $1.2 billion in a round led by Dragoneer and Lightspeed, pushing its valuation to approximately $18 billion, up from $14 billion when it last raised in June 2025. Helsing builds AI systems for Western defence forces, including signal intelligence software, autonomous system support, and battlefield awareness tools. The company was backed in its previous round by Spotify founder Daniel Ek and has been expanding across European defence programmes, including German armed forces contracts worth up to 1.46 billion euros over seven years.

Helsing positions itself as a sovereign European alternative to US defence AI providers such as Anduril, with roughly 80 per cent of its ownership remaining in European hands. European defence budgets have risen sharply since 2022, and AI-powered systems are now central to procurement planning across NATO member states. The near-$4 billion valuation jump in under a year illustrates the premium investors are placing on AI in high-stakes, regulated sectors. The company is preparing for an eventual public offering, though no timeline has been confirmed.

Israeli cybersecurity startup Frame Security has emerged from stealth with $50 million in funding led by Index Ventures, building a platform to protect organisations from AI-generated social engineering attacks. Frame was founded by veterans of Israel’s elite cyber Unit 8200 and includes Wiz alumni in its leadership. The company says it already has tens of enterprise customers, including Louis Dreyfus and Rockefeller Capital Management. Its platform analyses communication patterns in real time to detect AI-generated impersonation, rather than relying on known attack signatures that adversaries can update.

The core problem Frame addresses has got materially worse in the past 18 months. Generative AI has made convincing fake voices, synthetic video calls, and highly personalised phishing messages cheap to produce. For UK small and medium businesses, AI-powered CEO fraud and credential theft attempts have increased substantially, and many smaller organisations lack the infrastructure to detect sophisticated synthetic impersonation. Frame Security represents the growing market response to an attack surface that generative AI has widened, and its investor backing signals the problem is being taken seriously.

Here is everything else worth knowing from this morning’s AI news.

• ChatGPT adoption broadened sharply in Q1 2026. OpenAI’s own research shows fastest growth among users over 35 with more balanced gender usage, suggesting AI tools are moving well beyond the early-adopter demographic. [11 May]
• Google Finance AI is now live across Europe. Google’s AI-powered Finance experience, including Gemini-driven analysis and full local language support, is rolling out this week across European markets including the UK. [11 May]
• OpenAI formally launched its DeployCo subsidiary. Announced as a business built to bring frontier AI into production for enterprise clients. Today’s Decoder reporting adds depth on how the Palantir-style strategy is intended to work. [11 May]
• Cowboy Space raises $275 million for orbital data centres. Robinhood co-founder Baiju Bhatt’s startup closed a Series B led by Index Ventures to build AI compute infrastructure in orbit, addressing an infrastructure gap AI demand is driving. [11 May]
• Digg relaunches as an AI news aggregator. The once-popular bookmarking platform is back, using AI to surface high-signal news from influential voices. Currently in beta with invites open to subscribers. [11 May]
• State-linked hackers used AI to find a major software flaw. Google disclosed that threat actors used AI tools to identify a significant software vulnerability before it was patched, adding urgency to debates about AI-assisted cyberattack capabilities. [11 May]
• ChatGPT faces lawsuit over Florida State University shooting. OpenAI is named in legal action alleging that ChatGPT conversations contributed to the attack. The case raises AI liability questions that courts have not yet definitively addressed. [12 May]
• Interfaze publishes new high-accuracy model architecture. A startup called Interfaze has detailed a model architecture it claims delivers high accuracy at scale, with the announcement surfacing on the Hacker News front page this morning. [12 May]

This is a daily news update for informational purposes only. AI products and policies change rapidly. Verify details directly with providers before making decisions. Nothing here is financial or legal advice.

AI Daily is Cristoniq’s daily guide to developments in artificial intelligence, published every morning.