Safran press release, Paris 26 July 2011:
At the date of purchase, L-1 Identity Solutions, Inc., had never made a profit. Hardly surprising. The company was a ragbag of failed biometrics businesses, including Visionics Corp., Identix, Inc., and Viisage.
Safran completes the acquisition of L-1 Identity Solutions Becomes world leader in biometric identity solutions
After completing all required approval procedures, Safran (NYSE Euronext Paris: SAF) today announced that it has finalized the acquisition of L-1 Identity Solutions, Inc., a leading identity management solutions provider in the United States, for a total cash amount of $1.09 billion ($12 per share), which was originally announced in the press release on September 20, 2010. Following this transaction, Safran becomes the world leader in biometric identity solutions ...
L-1 will join Safran’s existing security business, operating as Morpho, and will be renamed MorphoTrust. The new company will be partly managed as a proxy structure, thus providing appropriate protection for U.S. national security ...
Jean-Paul Herteman, Chairman and CEO of Safran, said: "We are delighted to have finalized this transaction, which is perfectly aligned with the Group’s development strategy in the security business..."
Identix is particularly well known in the UK. In 2004, the UK Passport Service conducted a year-long trial of biometrics which proved that they are not reliable enough for use in passports, ID cards, residence permits, visas, driving licences and the like, please see cribsheet below. The trial was carried out using Identix products (Appendix C, p.254ff).
"$1.09 billion" may seem like a very precise number. It isn't. Unmentioned in the press release above, Safran took on about $500 million of L-1's debt in addition to buying the company. Safran's shareholders' initial stake is therefore a lot higher than $1.09 billion, please see for example this 16 May 2011 Bloomberg article:
And for that, Safran doesn't even get unfettered control. There's a "proxy structure" in there "providing appropriate protection for U.S. national security". Pleading national security, Safran's US Federal and State contracts could be switched to the all-American 3M Cogent, leaving Safran with nothing to show for $1.58 billion.
Safran, a Paris-based maker of airplane engines for Airbus SAS and Boeing Co., agreed to buy L-1 for $12 a share, or 48 percent more than L-1’s 20-day trading average before it was first reported July 15 that Safran was considering a purchase of L-1. The offer is valued at $1.58 billion including net debt.
You can see why L-1 would be pleased with this deal. It's not obvious what's in it for Safran.
This isn't the first time that shareholders and equity analysts will have had qualms about Safran's venture into biometrics.
On 7 October 2009, when their subsidiary Morpho was still known as "Sagem Sécurité", Safran issued this press release in Paris:
How did Safran/Sagem Sécurité/Morpho get this contract with IBM?
Sagem Sécurité chosen by IBM to support United Kingdom’s National Identity Assurance Service (NIAS)
Sagem Sécurité (Safran group) has signed a contract with IBM to supply and maintain a biometric management solution for British travel and identity documents, on behalf of the British Home Office’s Identity and Passport Service (IPS). The project is a core element of the Government’s plans to upgrade to biometric passports and enhance the security of the UK border.
Sagem Sécurité will provide multibiometric facial and fingerprint recognition technology that was assessed for speed, accuracy and cost in competitive trials developed and run by IBM, using in excess of 10 million images. The technology will enable IBM to help IPS and the UK Border Agency to deliver the next generation of secure and reliable identity documents to British citizens, residents and people requesting asylum, while minimising the risk of fraud ...
The answer is provided in a witness statement submitted by Mr Nicholas Swain in a case heard in the British courts, EA/2001/0081 (please see the entry for 20 July 2011). IBM organised a demonstration of biometric capability for the UK Home Office. Mr Swain is a Commercial Director at IBM and he says:
In 2008 and 2009 IBM had no particular expertise in biometrics. They have remedied that thanks to Safran, who gave them "business-critical intellectual property ... representing the results of major investment in software research and development".
10. As part of IBM's bid, during late 2008 and early 2009, IBM carried out a series of tests with specialist biometric software providers who were bidding to be part of ... IBM's solution for the NBIS project as part of the Demonstration ...
11. IBM negotiated the commercial arrangements with each of the biometric service providers, including Non-Disclosure Agreements (NDAs) to protect their commercially sensitive information (see further below). Six suppliers participated in the Demonstration ...
14. Thereafter, on 1 May 2009, IBM signed a contract to provide NBIS with the Home Office and, shortly afterwards, entered a sub-contract with Sagem (now Morpho), one of the suppliers who participated in the Demonstration, to provide the specialist biometric software needed for NBIS. In August 2010 this contract was revised and the programme was re-named IABS ...
18. All of the suppliers involved in the Demonstration made significant investment in time and provided IBM with more details about their products performance than is generally available. The information provided included business-critical intellectual property of the suppliers, representing the results of major investment in software research and development ...
IBM have played a blinder. They won a £265 million contract from the British government. And they acquired the fruits of several decades of Safran's R&D. All in return for a piece of paper, an NDA. You can see why IBM would be pleased with this deal. It's not obvious what's in it for Safran.
Safran's products can be tested without handing over the crown jewels. IBM and the Home Office only need to know whether Safran's products work, not how they work.
The directors of Safran gave the shareholders' intellectual property to IBM and they gave $1.6 billion of shareholders' money to L-1. What did they give to get their contract with UIDAI, the organisation responsible for issuing electronic IDs to 1.2 billion Indians?
Cribsheet – the failure of biometrics
Using L-1/Identix biometrics technology, the Home Office conducted trials of face recognition, fingerprinting and iris scanning back in 2004. The report on the trial was published by Atos Origin in May 2005 and even after many months of massage the figures still demonstrated failure.
10% of able-bodied participants in the trial couldn't register their irisprints in the first place, and that figure rises to 39% for the disabled participants. These people would quite simply not exist if public services only recognised people by their irisprints.
Face recognition biometrics failed with 31% of the able-bodied participants and 52% of the disabled. We would all do better to toss an unbiased coin than to rely on face recognition, a technology with an uninterrupted history of failure.
Which leaves us with fingerprints.
Understand that we're not talking here about traditional fingerprinting. The technology trusted by law enforcers worldwide for over a century now. Rolled prints. Taken using ink. By a police expert. Acceptable as evidence in a court of law. A technology so accurate that when there's a disagreement independent experts are flown in to resolve the matter.
No, we're talking instead about a modern, cheap, clean, quick technology, no expert required, a sort of glorified photocopying process, utterly unreliable, with a 19 or 20 percent failure rate. 19% for the able-bodied and 20% for the disabled. A technology that doesn't work well with old people, manual labourers, people from East Asia and women (p.34).
So much for L-1's biometrics technology. No-one is going to fly in independent experts from abroad to investigate 19 or 20 percent of all disputed matches and non-matches. Flat print fingerprinting, to put it loosely, doesn't work.
If the right to public services or the right to work or the right to vote or the right to a pension or the right to get married or the right to live in your municipal area or the right to travel beyond it – the right to cross an invisible eBorder – ever depend on flat print fingerprinting, then 19 or 20 percent of people legitimately entitled to those benefits will be wrongly denied them.
Do Morpho's other biometrics products work any better than L-1's?
So far, the public has not been told. Not in the UK, not in France, nowhere. Public money – your money and mine – is being invested by the UK Home Office and by Interior Ministries around the world, with no justification given.
Given that the only report on the reliability of biometrics published to date by the UK government demonstrates that the technology doesn't work, we need to see some independent and academically scrupulous evidence that our money isn't being wasted.
For all we know, the belief in the reliability of today's mass consumer biometrics is as foolish as the belief in astrology.
As Professor Ross Anderson, the king of IT security engineering, points out, the banks don't trust mass consumer biometrics technology. Otherwise they'd use it. So why does the government trust this technology? And why should we?
With no answers forthcoming, for all we know our money is being wasted on snake oil.