News briefing 27 Jan-14 Feb 2025

Autonomy in weapons systems, and AI in security and defence

• International: Google Lifts a Ban on Using Its AI for Weapons and Surveillance: This piece by Wired reports that Alphabet, Google’s parent company, updated its principles on applications of artificial intelligence to now allow its use for surveillance and weapon technologies, which it had previously pledged ‘not to pursue.’ It first published the AI principles in 2018 as ‘concrete standards that will actively govern our research and product development’ to acknowledge that ways in which AI ‘is developed and used will have a significant impact on society for many years to come.’ This rollback has been justified by the company through a blogpost to say that there have been advancements in the field of AI since 2018, and that it believes ‘democracies should lead in AI development, guided by core values like freedom, equality, and respect for human rights.’ Its new policy and principles on AI claims to address ‘evolving complexities and risks’ by ‘Implementing appropriate human oversight, due diligence, and feedback mechanisms’ as well as ‘rigorous design, testing, monitoring, and safeguards to mitigate unintended or harmful outcomes and avoid unfair bias.’ In another piece by Wired (originally in Italian), Dr Elke Schwarz highlighted the nexus between militaries and the tech industry. See here for more details from our previous news briefing. 
• International: Closer than ever: It is now 89 seconds to midnight: The Bulletin of Atomic Scientists has moved the Doomsday Clock to 89 seconds from midnight, closest it has ever been to catastrophe. One of the reasons for this, as mentioned in its post, is that ‘systems that incorporate artificial intelligence in military targeting have been used’ while countries are racing to ‘integrate artificial intelligence into their militaries.’ In another post, the Bulletin expresses concern on the integration of AI into decision support systems. 
• International: Autonomy and Accuracy in Weapon Systems: The Prospects of Artificial Intelligence: This comprehensive article on AI in weapon systems, autonomy, and accuracy by Professor Ernesto Damiani from the Department of Computer Science at Università Statale di Milano shares the view that AI has been integrated into weapon systems for ‘over three decades’, with supervised learning models playing a key role in autonomous decision-making. The evolution of AI, particularly with the emergence of Large Language Models (LLMs), has significantly impacted military applications, raising concerns about accuracy, autonomy, cybersecurity, and legal implications.
• International: UNSW awarded contract to develop framework for measuring ethical readiness of autonomous systems: This collaboration between UNSW Sydney and Arizona State University, funded by the Defense Advanced Research Projects Agency of the United States Autonomy Standards and Ideals with Military Operational Values (ASIMOV) program, aims to develop ethical benchmarks for autonomous weapon systems. Led by Professor Deane-Peter Baker, the project on Adversarial Generative Ethics for Autonomous Weapon Systems, aims to ‘break new ground by establishing a quantitative approach to measuring the ethical difficulty of future autonomy use-cases and the ability of autonomous systems to perform in those use-cases within the context of military operational values.’ 
• International: Every human life deserves safety and dignity: A call to make international humanitarian law a political priority: This post, the ICRC shares remarks delivered by Mirjana Spoljaric, president of the International Committee of the Red Cross (ICRC), at an event convened by the president of the UN General Assembly, Philemon Yang, with support from the ICRC, about the importance of upholding international humanitarian law. Spoljaric, in her remarks, calls for greater compliance with IHL, and urges on the need to ‘think about how the rules of war apply to new technologies, and consider the challenges that large-scale conflicts might bring.’ 
• International: Paris AI summit sidelines AI safety in favour of drumming up business: This video piece from France24 highlights comments from Dr Stuart Russel, professor of computer science at UC Berkeley, on the Global Artificial Intelligence Summit in Paris. He said that ‘the focus of the summit seems to be very much on the commercial opportunities.’ He warns that ‘AI systems, as they become more capable, pose the risk that we will not be able to control them.’ 
• International: SIPRI publication: Dilemmas in the policy debate on autonomous weapon systems: Dr Alexander Blanchard, Dr Vincent Boulanin, Laura Bruun and Netta Goussac at SIPRI write on the diversification of platforms to discuss autonomous weapon systems, and the possible ‘trade offs’ that entail with it. Acknowledging the frustration ‘at the slow rate of progress in the GGE on LAWS and its limited participation and scope’, the piece calls for ‘policymakers considering the various calls to continue diversifying the process to weigh up the benefits and possible trade-offs involved.’ 
• India: India’s Approach to AI in Military Domain & Emerging Technologies in Areas of Lethal Autonomous Weapon Systems (Laws): Manohar Parrikar Institute for Defence Studies and Analyses, in collaboration with the Indian ministries of external affairs and defence organised a three days workshop on autonomous weapon systems. The event attendance was through invite only, and sought to address the necessity to balance benefits of AI integration in the military ‘against the challenges of legal accountability, compliance with international humanitarian law, and the ethical considerations of delegating life-and-death decisions to machines.’ 

AI, algorithms and autonomy in the wider world

• UK: UK to become first country to criminalise AI child abuse tools: The UK will be the first country to criminalize AI-generated child sexual abuse content. New laws will make it illegal to create, possess, or distribute AI tools that generate such material, with penalties of up to five years in prison. It will also criminalize AI-generated “paedophile manuals” and websites facilitating child abuse, punishable by up to 10 years. Home Secretary Yvette Cooper warns that AI is accelerating child exploitation, urging global action. The Internet Watch Foundation had noted an increase in AI-generated child abuse images, emphasizing the need for stricter regulations.
• EU: The EU’s AI bans come with big loopholes for police: EU’s AI Act came into effect early February this year, which will ban the use of AI for ‘unacceptable uses’ including predictive policing, reports Politico. While there is great anticipation to see how these provisions are implemented, rights groups have raised concerns regarding rampant uses of AI by migration authorities. 
• International: AI ‘godfather’ predicts another revolution in the tech in next five years: The Guardian reports on a panel at the Queen Elizabeth prize for engineering, with former recipients of the award, Yoshua Bengio, Yann LeCun, and Bill Dally. Yann LeCun, the chief AI scientist at Mark Zuckerberg’s Meta, spoke on how present AI systems were limited by their lack of perceiving the real world, and hence are operating by ‘manipulating language’ at present. Yoshua Bengio said at the same panel that AI regulation is crucial to address risks posed by it. 

Facial recognition, biometric identification, surveillance

• Palestine: ‘The Palestine Laboratory’ exposes Israel’s export of unique systems of control and surveillance technology around the world: Al Jazeera English has released a two part film titled “The Palestine Laboratory” by journalist Antony Loewenstein which ‘investigates how Israeli weapons and surveillance technologies are first used to control and repress Palestinians in Palestine before being sold all over the world.’ Both episodes are now released on the agency’s youtube channel. 
• Belize: Belize to install 1,000 surveillance cameras with facial recognition from March: This piece by Biometric Update reports on police authorities in Belize preparing to install around 1000 surveillance cameras in Belize city, under the pretext of ‘enhancing public security and safety.’ The system would reportedly have ‘motion video detection capabilities, will be able to track unusual movements and send an alarm to alert the police of such incidents.’ 
• Brazil: Brazil bans TFH from offering compensation for biometric data: Biometric Update reports that Brazil’s ‘National Data Protection Authority (ANPD) issued a preventive measure against the venture, formerly known as Worldcoin’ effectively banning their project offering cryptocurrency against consent for biometric data collection from people. See here for previous news briefing coverage of Brazil’s data privacy regulator.
• International: Border projects see criticism on data sharing transparency: Biometric Update has raised concerns that Governments worldwide are expanding biometric technologies for border security, raising concerns about data privacy and transparency. Switzerland plans to extend facial recognition at airports, potentially bypassing consent laws. In North America, a new U.S.-Canada agreement increases biometric data sharing without external oversight, sparking criticism. Meanwhile, the EU’s Odysseus project, which uses AI to assess travel risks, faces scrutiny for lacking transparency. Critics argue that these developments could undermine privacy and data protection standards globally.
• Finland: Finland inches towards granting wider access to biometric data to police: Finland is considering expanding police access to biometric data from passports and ID cards to investigate serious crimes and prevent threats. The proposal, currently under public review, would allow law enforcement and intelligence agencies to use fingerprints and facial images under strict regulations. The bill is expected to be submitted to Parliament in August this year.
• US: Clearview fights to give away 23% of facial recognition business: A US Judge is reviewing a proposed settlement in a biometric privacy lawsuit against Clearview AI, which would grant plaintiffs a 23% stake in the company. Critics argue this contradicts efforts to shut Clearview down.
• US: Judge throws out Clearview facial recognition match in US murder case: This story highlights the lack of training for law enforcement in using facial recognition. It has also put the spotlight on other law enforcement agencies across the US who are using Clearview’s technology – despite controversies. See here for previous news briefing coverage of Clearview. 
• International: North Korea tightens surveillance of citizens for possession of outside info: UN human rights report: A UN report reveals that North Korea has intensified surveillance and repression against citizens. The report can be found here.