Article content
Successfully implementing artificial intelligence across all organisations, including those involved with defence and security, will require more than just investments in software and technology. Instead, shifts in organisational structure, internal culture, and strategy execution are necessary, according to Bertrand Rondepierre, General Manager of the French Agency for Artificial Intelligence in Defence (AMIAD) and a keynote speaker at the Tallinn Digital Summit.
Speaking at last month’s summit, Rondepierre said that he had learned over the past decade, during which he worked in both the private and public sectors, that initial attitudes toward artificial intelligence had not allowed organisations to capitalise on its full potential. Instead of focusing on acquiring platforms, he urged organisations to think much bigger about AI.
There is a need to change organisations, processes, and systems to become AI-native, which will require investing in the people, culture, and infrastructure that enable AI to function effectively, he said.
Winning the conflicts of the future
During his talk at the summit, held at the National Library of Estonia in mid-October, Rondepierre noted that the way AI is used on the battlefield is evolving, from utilising high-cost, high-tech hardware to more innovative, mass-produced, lower-cost systems. The question facing the defence sector is whether their investments in hardware or in computational power will win the conflicts of the future. Rondepierre agreed that investments in GPU and compute infrastructure are of importance, but he stressed that France is undertaking a more comprehensive approach.
First, he said that France is focused on hiring the right talent, defining clear missions for its professionals, and compensating them fairly. France is also investing in the GPUs, storage, and interconnects that are needed to scale AI and connect systems, data, and people, he said.
France is also intensifying its collaborations with industry. “We can’t do everything alone, and Europe’s strength, and challenge, lies in its many diverse players,” he said. He acknowledged that such partnerships are complex because of procurement, bureaucracy, and digital infrastructure, but he maintained that these were prime areas for improvement in Europe.
The next frontier
In his keynote, Rondepierre cautioned that structures built to implement strategy can, in fact, be impediments to AI adoption. He said that cybersecurity should not paralyse operations and that processes across the board should be simplified.
Rondepierre added that within the AMIAD, which was set up last year and currently employs 150 employees, an “ambitious product guideline” is being created, one that will prioritise agentic AI and robotics. “I believe AI’s next frontier is in the real world: in robotics, on the battlefield, and in physical systems, not just in cyberspace,” said Rondepierre. By taking such an approach, Rondepierre concluded, France should be able to “deliver value at scale” for its military personnel.
Code, conflict, and cooperation
Rondepierre had the opportunity to elaborate on the themes touched on in his talk in the first panel of the conference. Entitled, “Code, Conflict and Cooperation: When AI Becomes Grand Power Warfare,” the panel included Rondepierre; Shon Manasco, the head of international defence at Denver, Colorado-based Palantir Technologies, and Sten Tamkivi, an Estonian entrepreneur and a partner at Plural, a venture capital firm based in Tallinn and London.
The panel was moderated by Kristen Davis, founder and CEO of CinqC, a Paris-based consultancy. She opened by noting the impact of AI on defence and security, from the way borders are defended to how weapon systems are procured, and asked the panellists about the best ways to “master the tools of this new arms race.” Should Europe outsource military capacity to partners, pursue sovereign strategies, or build entirely new doctrines, she asked.
A defence tech renaissance
The panellists noted that Russia’s full-scale invasion of Ukraine has created a renaissance in defence tech. Tamkivi noted that Plural has backed several companies in the space, including Helsing, an AI defence unicorn, and Labrys, based in London, which offers a software platform for coordinating global security teams, civilian and defence networks, in both peace and conflict.
Palantir’s Monasco noted that the company’s software has been in use in Ukraine since the start of the invasion. The Palantir platform “fuses data from sensors in orbit, in the air, and on the ground in real time.” This gives operators a clear understanding of what’s happening and helps them make better decisions about deploying weapons and defending their territory, he said.
Rondepierre noted that AMIAD has been growing rapidly to meet its mission and aims to have 300 staff members by the end of 2026. He said that the agency is now working to scale internally and with its industrial partners, while maintaining its sovereignty. “We need to be able to operate and manage our own systems so that if, for any reason, we lose access to external partners, we can still function independently,” said Rondespierre at Tallinn Digital Summit. “Without in-house engineers and technical capabilities, we’d be completely reliant on others,” he said, “and that’s not sustainable.”
Investing in talent and scale
Indeed, part of scaling means hiring new talent. And all the panellists underscored this need. According to Rondepierre, gaining new talent means paying them well, which should involve fair compensation and salaries that are competitive with those available in the private sector.
Talent should also feel inspired, and Rondepierre noted that the Russo-Ukrainian War has been such a catalyst.
“The war … has awakened something in the tech community, a realisation that the world is dangerous and that we all have a role in strengthening defence and resilience,” he said. “Many talented people want to contribute to that mission, not just in startups but in the government.”
Tamkivi said he was optimistic about the European talent pool. He noted that Europe has more top 200 universities than the US, around 100,000 advanced AI researchers compared to about 80,000 in the US, which should make it more competitive. To attract and retain talent, companies must offer qualified mentors and challenges that are worth solving. “If you offer those, talent will come and stay,” he said.
Monasco noted that a third of Palantir’s workforce is based in Europe. “When NATO adopted our software, many engineers from Europe who were working in the US said, ‘I want to go home. I want to help strengthen security in my own country,” he said. “That was really powerful to see.”
Another challenge, Tamkivi said, is scaling that talent and growing companies to achieve a global scope. Achieving that will mean improving rules around stock option taxation, startup visas, and “other barriers that slow down innovation.”
Rondepierre noted that governments, too, need to be good clients to industry. “The best thing we can do for startups isn’t to subsidise them,” he said. It’s to buy their products.” He also noted that the process of purchasing those products could be streamlined, thereby overcoming months of negotiation, integration, and contracting. By that time, the business could already be lost to the US market.
A source of prosperity
At the end of the discussion, Davis asked the panellists for metrics of success for the use of AI in defence in Europe. Monasco said that the health of the ecosystem could be measured in terms of the number of European companies delivering capabilities and winning defence contracts across NATO. He also noted that preparedness could be measured by how many companies were providing “tangible, deployable tools that strengthen our collective defence.”
Tamkivi said one metric for success could be a shift from seeing defence as an expense to a revenue source. “I’d like to see Europe move from talking about defence spending to defence income,” he said.
Should Europe develop the best defence technologies, those could be exported, generating revenue and strengthening Europe’s economies. “Europe’s tech for drone warfare and battlefield automation is already among the best in the world because it’s being tested here, in real conditions.” He said, “Five years from now, those solutions could be our biggest exports.” Europe could thus change defence from a cost centre into a source of prosperity, said Tamkivi at Tallinn Digital Summit.
Rondepierre suggested that his job would have been successful if an AI agency were no longer necessary in France in a decade. “When AI is fully integrated into daily operations and thinking, that’s when we’ll know we’ve succeeded,” he said.
